ASSISTIVE TECHNOLOGY TO ENHANCE WRITTEN EXPRESSION OF STRUGGLING WRITERS IN ELEMENTARY SCHOOL: A TABLET-BASED LITERACY INTERVENTION PROJECT by: HEATHER STACE-SMITH A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Master of Arts in Educational Studies, Special Education We accept this thesis as conforming to the required standard …………………………………………………….. Dr. Kenneth Pudlas, Ed.D. Thesis Supervisor …………………………………………………….. Dr. Katrina Korb, Ph.D., Second Reader …………………………………………………….. Dr. Julie Corkett, Ph.D., External Examiner TRINITY WESTERN UNIVERSITY December, 2017 © Heather Stace-Smith ASSISTIVE TECHNOLOGY TO ENHANCE WRITING ii ABSTRACT Assistive technology has been extensively used as method of improving learning for struggling students, despite the lack of empirical research to support this practice (Bebell & Pedulla, 2015; Cumming, Strnadová & Singh, 2014; Maor, Currie & Drewry, 2011). In an effort to discover effective strategies for struggling writers, the researcher investigated the effectiveness of the iPad application Clicker Docs, in combination with built-in tablet accessibility features as an intervention tool to improve writing for struggling writers. Using a switching replications quasiexperimental design, this study investigated the effectiveness of the application Clicker Docs and tablet accessibility features as a 6 week alternating intervention tool for improving writing. Aspects of writing included writing quality, as measured by student writing samples assessed with a teacher developed rubric, writing output, as measured by number of words per writing sample, and attitudes of struggling writers, as measured by the Writing Attitude Survey (Kear, Coffman, McKenna, & Ambrosio, 2000). Two groups of 11 students from grades 2-7 in a small rural school in B.C., who were identified with a disability or as a struggling writer, alternated participation in this intervention program that included two 25 minute intervention sessions per week. During the intervention program, every student was taught how to use accessibility features as well as how to use the application Clicker Docs on their own personal device. It was hypothesized that overall the writing of the students would improve following the iPad intervention. A mixed 2x2 repeated measures analysis of covariance (ANCOVA) with pre-test scores as covariate was used to analyze the results for each specific research question. Results showed a large significant effect of the iPad application Clicker Docs and accessibility features on writing quality at Post-test 2. On average, those in the iPad intervention group demonstrated better writing quality than those in the control group, when factoring in pre-test scores. In ASSISTIVE TECHNOLOGY TO ENHANCE WRITING iii addition, a medium significant effect of the iPad application Clicker Docs and accessibility features was found for writing output. Contrary to researcher hypotheses, on average, those in the iPad intervention group wrote less overall than those in the control group, when factoring in pre-test scores. No effect of the iPad application Clicker Docs and accessibility features was found for attitude towards writing. Keywords: assistive technology, tablet computer, iPad, struggling writers, learning disabilities, writing quality, writing output, writing attitude, intervention, switching replications design, ASSISTIVE TECHNOLOGY TO ENHANCE WRITING iv TABLE OF CONTENTS ABSTRACT...................................................................................................................................................... ii TABLE OF CONTENTS .......................................................................................................................... iv LIST OF TABLES ............................................................................................................................................. vi LIST OF FIGURES .......................................................................................................................................... vii AKNOWLEDGEMENTS ................................................................................................................................ viii CHAPTER 1: INTRODUCTION AND BACKGROUND ........................................................................................ 1 Introduction .............................................................................................................................................. 1 Background ............................................................................................................................................... 3 Project Description.................................................................................................................................... 5 Project purpose and objectives. ........................................................................................................... 5 Relevance and significance. .................................................................................................................. 5 CHAPTER 2: LITERATURE REVIEW ................................................................................................................. 8 Historical Context of Technology to support Writing ............................................................................... 8 Assistive Technology to Support Learning ................................................................................................ 9 Mobile Devices to Support Learning ....................................................................................................... 14 Potential of Tablet Computers to support Learning for Children with Disabilities ................................ 16 The Use of Tablet Computers to Support Writing .................................................................................. 21 Summary of Literature Review ............................................................................................................... 25 Research Problem ................................................................................................................................... 27 CHAPTER 3: RESEARCH DESIGN, METHOD, AND PROCEDURES.................................................................. 29 Research Design ...................................................................................................................................... 29 Participants ............................................................................................................................................. 31 Instruments ............................................................................................................................................. 34 Research Materials ................................................................................................................................. 35 Procedure................................................................................................................................................ 37 Data Analysis ........................................................................................................................................... 39 Ethics ........................................................................................................................................................... 40 Approval and consent. ........................................................................................................................ 40 Anonymity and confidentiality............................................................................................................ 40 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING v Potential conflict of interest. .............................................................................................................. 40 Scientific Quality: Validity and Reliability ............................................................................................... 41 CHAPTER 4: RESULTS................................................................................................................................... 43 Question 1: Writing Quality .................................................................................................................... 43 Question 2: Writing Output .................................................................................................................... 46 Question 3: Attitude towards Writing .................................................................................................... 49 CHAPTER 5: DISCUSSION ............................................................................................................................. 51 Question 1: Writing Quality .................................................................................................................... 51 Question 2: Writing Output .................................................................................................................... 53 Question 3: Attitude towards Writing .................................................................................................... 55 CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS ............................................................................. 57 Summary ................................................................................................................................................. 57 Implications of Research ......................................................................................................................... 57 Limitations .............................................................................................................................................. 59 Considerations for Future Research ....................................................................................................... 60 Concluding Thoughts .............................................................................................................................. 61 REFERENCES ................................................................................................................................................ 62 APPENDIX A: Participant Raw Data ............................................................................................................. 69 APPENDIX B: Writing Quality Assessment Instructions .............................................................................. 70 APPENDIX C: Writing Quality Assessment Grade Level Rubrics ................................................................. 71 APPENDIX D: Writing Attitude Survey Sample Page ................................................................................... 77 APPENDIX E: Screenshots of the Clicker Docs Application by Crick Software ............................................ 78 APPENDIX F: Assistive Technology Intervention Lesson Plans.................................................................... 80 APPENDIX G: Planning for Writing Assessment Page and Writing Paper ................................................... 90 APPENDIX H : SPSS Data Analysis for Question 1 ....................................................................................... 92 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING vi LIST OF TABLES Table 1. Gender of participants………………………………………………………………..…33 Table 2. Scores on screener writing tests………………………...………………………………33 Table 3. Diagnosed disabilities of participants…………………………………………………..34 Table 4. Assumptions tested for each writing assessment dependent variable…………………..43 Table 5. Interaction effects of the independent variables on each dependent variable…………..44 Table 6. Mean scores on writing assessment post-tests……………………………………….....46 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING vii LIST OF FIGURES Figure 1. Hypothesized writing assessment outcomes at each testing period…………………...28 Figure 2. Switching Replications Quasi-Experimental Design………………………………….30 Figure 3. Mean writing quality composite scores by group……………………………………..45 Figure 4. Mean writing output composite scores by group……………………………………...47 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING viii AKNOWLEDGEMENTS I am so very grateful to the many individuals without whom this thesis would not have been possible. Many thanks to my advisor and mentor Dr. Ken Pudlas for encouraging me to embark upon this journey, and for continually affirming me in my ability to do so. Thank you also to my second reader Dr. Katrina Korb, who did so much more than second readers are required to do. I appreciate your countless hours of continual teaching and support. You enabled me to achieve a level of understanding of statistics and research design that I would have never imagined possible. Thanks also to my amazing cohort for providing much needed emotional support, and for giving me the confidence that I didn’t have for myself. I would like to thank the members of my school community for allowing this research project to come into existence. I would like to thank each of my research participants for their willingness to learn and to support learning. Thanks also to my fellow staff members for your unending support and your willingness to be continually flexible. I would like to thank my family for all of the sacrifices made. I would like to thank my children for cheering me on, and for their continual prayers for “Mommy to get her work done” and for “Mommy to graduate”. I would like to express my deepest gratitude to my husband Rob, whose constant and unwavering support is too monumental to express in words. Finally I would like to thank my Heavenly father for the honor of being called to this Graduate program. Thank you for allowing me to call upon your strength to accomplish things that would have otherwise been impossible. For with God nothing will be impossible- Luke 1:37 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 1 CHAPTER 1: INTRODUCTION AND BACKGROUND Introduction Competency in writing skills is foundational for many areas in life (Asaro-Saddler, Knox, Meredith, & Akhmedjanova, 2015; Wollak & Koppenhaver, 2011). Therefore, building literacy skills in the areas of writing is given tremendous importance in schools. According to the Education Quality Accountability Office (2013, as cited in Dunn, 2015), approximately 40% of Canadian students are not able to write at a basic level of ability. Defining appropriate instructional strategies to overcome barriers to writing then becomes a crucial task (Sessions, Kang, & Womack, 2016). The use of assistive technology is a strategy that is often promoted to help students improve in writing skills (Flanagan, Bouck, & Richardson, 2013). It seems to be a widely held assumption that digital technology can serve as a means for enhancing student performance in school (Cumming et al., 2014; Peterson-Karlan, 2011; Suhr, Hernandez, Grimes, & Warschauer, 2010). Unfortunately there is a lack of research-based evidence to support this assumption (Cumming et al., 2014). However, there is some research to support technology as a tool to increase academic achievement (Larabee, Burns, & McComas, 2014). Research to support technology as a tool for improvement in writing specifically is beginning to emerge (Cullen, Richards, & Frank, 2009; Peterson-Karlan, 2011). Historically, assistive technology has been promoted as a medium to help break down barriers created by learning challenges, especially for children with learning disabilities (Adebisi, Liman & Longpoe, 2015). Assistive technology are the equipment, devices, services, processes, systems, and adaptations made to the learning environment that support and enable their functions, used by persons with special needs (Erdem, 2017). An assistive technology tool is any item that is used to maintain or improve the functioning of a child with a disability (Adebisi et ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 2 al., 2015). The purpose of assistive technology devices is to help individuals with disabilities function better in everyday life as well as attain a higher quality of life (Erdem, 2017). Children with learning disabilities are especially in need of intervention as they demonstrate decreased skills that do not improve as they progress in schooling under typical conditions of instruction (Peterson-Karlan, 2011). Assistive technology has been known to help students with special needs increase autonomy, develop independent thinking and problem-solving skills, maintain self-reliance, facilitate a sense of continuity in living conditions, and become more actively involved in their educational activities at home, schools and communities (Akpan & Beard, 2013). One type of technological tool that demonstrates potential in helping children with disabilities to overcome challenges in learning are tablet computers and their corresponding applications (Cumming et al., 2014). Tablet computers, commonly referred to as tablets, are wireless, portable devices that are generally smaller than laptop computers and larger than smartphones. A touch screen is commonly used to input information. These devices have a similar interface to smartphones with a wide variety of applications available to download. Tablet computers are showing much promise in enhancing literacy skills, especially for children with disabilities (Asaro-Saddler et al., 2015; Dunn, Barrio, & Hsiao, 2016; Johnson, 2013; Kagohara et al., 2013), and struggling or at-risk learners (Delacruz, 2014). They have many of the benefits of laptops and desktop computers for enhancing learning with the added benefit of built-in accessibility features that serve as assistive technology. Assistive technology can include features such as speech to text, word prediction, spell checker, grammar checker, speech synthesizers or e-readers, speech recognition, variable speech control, and educational software. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 3 The release of the Apple iPad in 2010 and other tablet computers shortly after, has allowed for increased assistive technology (Spooner, Kemp-lnman, Ahlgrim-Delzell, Wood & Ley Davis, 2015) that is more accessible as a result of built-in accessibility features (Maich & Hall, 2016). These features, combined with tablet computers’ portability, interactivity, and range of educational applications, allow for a personalized learning experience that can meet a range of diverse learning needs (Maich & Hall, 2016; Sessions et al., 2016). The Apple iPad specifically has been found to enhance student motivation and interest in writing (Cumming et al., 2014; Larabee et al., 2014; Musti-Rao, Lo & Plati, 2015; Sessions et al., 2016), problems often found with struggling writers. Unfortunately a search of the literature, using the search terms “tablet”, “tablet computer”, “iPad”, “mobile device”, “mobile learning”, “literacy”, “learning” and “writing”, entered into the search engines Academic Search complete, ERIC, E-Journals, psycARTICLES, psycINFO, Sage Journals, and Google Scholar reveals that there is currently a scarcity of research on using tablet computers such as the Apple iPad as a literacy intervention. Most of the available research focuses on improving reading and considerably less focuses on writing (Sessions et al., 2016). As the area of using tablets computers to support learning is a fairly new area of study, many of the studies are exploratory (Larabee et al., 2014). There are very few experimental studies involving tablet computers that directly investigate objective measures of student achievement, and most have methodological limitations (Bebell & Pedulla, 2015). Background The foundations for this study began in the school where the researcher is currently employed as both an inclusive teacher of elementary students, and as Special Education Coordinator. This school has a proportionally high percentage of students with disabilities and ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 4 learning challenges. Approximately 23% of the students in this school have individualized education plans. Thus, it has become increasingly difficult to meet the needs of all learners in their classrooms. Many students struggle with reading and writing and lack engagement in the learning process. Class-wide assessments performed at the beginning of the school year on students in grades 2-7 revealed low writing quality and skills for a high number of students. Teachers’ qualitative observations also revealed concerns with students’ writing ability and attitudes. Students were often observed to be frustrated or anxious at times involving writing, some students to the point of tears. Off-task behavior at writing time was also observed to be common. Many students were inhibited by the inability to spell words and would make comments such as “I’m just thinking of words I can spell before I can write” or “I can’t write that sentence because I don’t know how to spell it”. This seemed to result in very basic sentences with common vocabulary that the students felt they could spell. Getting started with writing was also a common problem with students frequently commenting “I don’t know what to write”. Students with fine motor challenges also struggled to physically write their ideas. The process of writing was so labour-intensive for some students that their ideas would become lost before they could get them on paper. As teachers were concerned about the number of students requiring intervention in the area of independent writing, a larger school-wide project aimed at improving literacy outcomes was undertaken with a provincial Ministry of Education resource program, whose mandate is to provide assistive technology for students with disabilities. This project resulted in a grant that enabled the school to acquire Apple iPad tablet computers for student use, as well as accessories and software recommended by a provincial teacher specialist for improving student writing. Training was given to all of the school staff by the provincial teacher specialist on using tablet ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 5 accessibility features as well as the provided software, and a unique opportunity for research was created. Project Description Project purpose and objectives. The purpose of this study is to experimentally examine the effectiveness of tablet computers as an intervention for improving struggling students’ writing output, writing quality, and attitude towards writing. This research focused on Clicker Docs, an application for iPad, combined with built-in tablet accessibility features as an intervention. The Clicker software family, used extensively as a desktop computer application, is shown to meet best practices for early literacy intervention (Parette, Hourcade, Dinelli & Boeckmann, 2009). It now has versions of the popular software available for mobile devices. It was hypothesized that overall the writing quality of the students would improve following the intervention. The objectives of this study include the following:  To determine the effect of assistive technology on writing quality among students with writing difficulties.  To determine the effect of assistive technology on writing output among students with writing difficulties.  To determine the effect of assistive technology on attitude towards writing among students with writing difficulties. Relevance and significance. In this specific school context, it is paramount to ensure the strategies being used to teach diverse learners are effective in helping them to learn, especially given the extent to which many ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 6 of the students struggle with writing. Discovering what supports are going to make a difference to change the trend of learning experiences and improve confidence is important for guiding future teaching strategies. It is hoped that such an intensive intervention program will be effective in increasing learning for all participants, especially those who have not yet experienced a degree of success with writing. Additionally, given the high cost of assistive technology, it is important to ensure any potential learning benefits to students justify the cost of using it as a tool more extensively in the future. Technology in the classroom is “inevitable” (Alnahdi, 2014, p.18) and “ubiquitous” (Spooner et al., 2015, p. 52). According to a 2011 Report by Simba Information, academic use of digital devices in the classroom is increasing with up to 75% of educators reporting that device technology including tablets, smartphones, eReaders, and MP3 players, is used by students in their districts for educational purposes in school (Raugust, 2011, as cited in Davis, Orr, Kong, & Lin, 2015). Consistent with the increase in mobile devices in classrooms, the number of instructional programs and educational applications have also continued to grow (Davis et al., 2015). A plethora of technologies are currently available, and new technologies are frequently developed. Hence, technology advances faster than it can be evaluated. Sung, Chang, and Liu (2016) note that there are limited studies that have addressed best practices in how to use mobile devices as well as the effectiveness of use. As previously mentioned, little research has been done on assistive technologies to support writing. Thus, more research is needed to determine what is effective, and how to use assistive technology effectively. As at least 40% of Canadian students are struggling writers (Dunn, 2015), it may be necessary to think outside the box to get children interested in, confident with, and progressing in their writing skills. Researchers have suggested that assistive technology could be one such tool effective in supporting writing for ASSISTIVE TECHNOLOGY TO ENHANCE WRITING students who are struggling (Adebesi, Liman, & Longpoe, 2015; Cullen et al., 2009, Mezei & Heller, 2012; Perelmutter, McGregor, & Gordon, 2017). Given the aforementioned dearth of recent research, contributing to scholarship by experimentally evaluating the effectiveness of assistive technology as an intervention for struggling writers may help to fill the gap in existing literature, as findings may be relevant to similar contexts. 7 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 8 CHAPTER 2: LITERATURE REVIEW Historical Context of Technology to support Writing There is a long historical basis of using technology to support writing in schools. Peterson-Karlan (2011) conducted a descriptive analysis of peer-reviewed literature examining the efficacy of technology to aid the compositional writing of students with learning disabilities. Based on selection criteria, 85 applied research studies ranging from 1984-2010 were reviewed. These studies were then grouped according to individual writing process areas (planning, transcription, editing, and revising) for analysis. Peterson-Karlan noted evidence-based practice criteria for number of studies and number of participants was not met in either of these four areas, meaning that there was not enough studies and enough participants to be able to properly analyze the quality of studies and effect sizes. Moreover, the trend for research in this area has declined significantly, with 65 of these studies being completed between 1984 and 2000, 13 completed between 2001 and 2005, and only 5 of the studies completed between 2005 and 2010. Therefore, the trend for advances in technology does not match the trend in research, meaning that much of the newer improvements in technology are not being evaluated. Peterson-Karlan (2011) was not able to confirm or disprove technology as an evidencebased practice due to the extreme paucity of research. However, he suggests there might be enough research to support technology as a promising practice. As the pace of advances in technology continue, much of the technology in these studies are outdated. These factors, taken together with the substantial declining trend in the number of applied research studies despite increases in technologies in schools, identifies a critical need for further research in this area. Similarly, Batorowicz, Missiuna, and Pollock (2012) investigated the outcome of using technology to support the written productivity of children with learning disabilities by ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 9 completing a review of 28 studies published between 1985 and 2010. Their findings were inconclusive overall. However, trends analyzed suggested a positive influence of some technology on children’s behavior and performance, thus giving some evidence to support the use of technology for writing of children with learning disabilities. However, it was noted that the level of this evidence is moderately small. It was found that technology may positively impact children’s attitudes, independence, and motivation for writing. Some technologies were seen to compensate for weaknesses and capitalize on strengths of students with learning disabilities. Technology was found to work better for low performers. The authors of the study note methodological limitations in most of the studies and note that most studies combine technology with teacher instruction. Hence, Batorowicz and colleagues likewise highlight the need for high-quality investigations with newer technologies. This is especially important given how long ago many of these studies were completed. Assistive Technology to Support Learning The impact of assistive technology features used on computers has been explored in the past. Maor and colleagues (2011) conducted a meta-analysis on empirical, peer-reviewed studies published from 2004-2009 that investigated the effect of software-based assistive technology on the reading, writing, spelling, or speech of children with disabilities. After an intensive search and selection process, only 15 studies were chosen. Most of the studies used an intervention with students who had physical or cognitive disabilities and used a pretest-post-test design with a control group. A few studies also used interviews or case-studies. Although some programs showed little to no difference as a result of the assistive technology intervention, the majority of studies found consistently improved outcomes. Almost all of the studies were able to demonstrate an increase of skill in the area tested. Limitations of this study included a wide ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 10 range of disabilities and technologies explored, some small sample sizes, and sometimes lack of control groups. As is the case with the meta-analysis completed by Peterson-Karlan (2011), much of the technology in these studies is now outdated. Erdem (2017) also completed a review of the use of assistive technology in the education of students with special needs. He reviewed publications in both the Journal of Special Education Technology and the International Journal of Special Education dated between 2010 and 2015, as well as a variety of other sources found in technology centres and by scanning the databases. Assistive technology was found to be implemented in the areas of “communication, reading, writing, mathematics, seeing and hearing skills, positioning-sitting and movement skills, social skills and making use of leisure time, daily life skills, organization and working skills, and computer skills” (Erdem, 2017, p. 130). The author found that various types of assistive technology are used in special education that overall generally have a positive effect. In a rigorous evidence-based systematic review and meta-analysis, Perelmutter and colleagues (2017) examined the effectiveness of assistive technology interventions for adolescents and adults with learning disabilities. They also examined the experiences of individuals with learning disabilities who were using assistive technology supports. A total of 38 quantitative group-design and single subject intervention studies, five survey studies, and 13 qualitative studies were examined following a thorough search of literature. Based on the review of qualitative studies, the authors found a number of students showed clear benefit as a result of assistive technology use. The importance of providing technical support was further noted. The review of quantitative studies allowed the researcher to draw conclusions about five different forms of assistive technology: text-to-speech including complex computer-based interventions with a primary text-to-speech component, speech-to-text, word processing including spell or ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 11 grammar check, multimedia and hypertext, and smart pens. Text-to-speech interventions, in which the computer reads out text to the user, were found to have a small positive effect. Studies using speech-to-text interventions, in which the user’s voice is translated into computer commands, were found to have positive outcomes. However, the studies were few in number as well as quite heterogeneous in research design and outcome variables. Therefore, no metaanalysis was conducted. Word processing applications that include spell-check and grammarcheck were found to have large beneficial effects. Multimedia interventions yielded strong positive effects. Smart pens, that is handheld devices with built-in scanning and character recognition features, were found to have statistically significant results, but with very small effect. The authors concluded assisted technology interventions are often helpful for adolescents and adults with learning disabilities. However, interventions need to be compared with each other before selecting interventions that are customized to the individual. This finding is of credit to the present research, in which the assistive technology features used by the participants can be customized through the settings on the application Clicker Docs. A limitation of this study is again the age of some of the studies included. The authors report that it is likely that many of the technologies have improved greatly since the time of the studies. Yet, given the trend towards positive outcomes in the past, advances in technology may lead to even greater outcomes (Perelmutter et al., 2017). . A study on teacher perceptions of assistive technology use has lent much insight into the perceived effectiveness in the classroom. Flanagan and colleagues (2013) examined the perceptions of middle school special education teachers of assistive technology used during literacy instruction with students with mild disabilities. A survey that explored the use, effectiveness, and factors impacting use or effectiveness of assistive technology in literacy ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 12 learning was sent to 166 school in one Midwestern state in the U.S.A. Fifty-one surveys were analyzed. As the survey included a variety of question types including single-selection, rating scales, multiple selection items, and open-ended responses; a variety of analyses were completed to analyze the data. Descriptive statistics, frequency distributions, cross-tabulations, and correlations revealed a number of themes. Results showed that even though teachers believed assistive technology to be an effective intervention for literacy, they use it minimally. Assistive technology was effective in that it made the education experience more equal through the accommodation to literacy curricula it provided. Moreover, previous positive experiences and knowledge determined the effectiveness of assistive technology. Barriers to using technology effectively were reported as cost, usability, and lack of training and experience. Therefore, a need for further training and professional development in using assistive technology devices was highlighted. This gives credence to the importance of the iPad training period given to staff in the present study. Overall, this study was rigorous and thorough. The limitations listed were slight. Yet, like so many of the other studies, the technology examined is not up to date, and mobile devices were not looked at specifically. Using a quantitative single group pre-test post-test design, Racicot (2016) investigated the effects of assistive technology incorporated in a multimedia writing support software program on the writing productivity and writer self-perception of 22 elementary students with mild to moderate delays. Students were first assessed using the Developing Writer’s Assessment and Adapted Writer Self-Perception Scale. An intense 8-week writing intervention consisting of two-45 minute sessions per week was delivered to all students. The Clicker 5 software and an interactive SMART board touch screen was used as the intervention technology. All of the students then completed a post-test using the same measures. Data were analyzed using a paired ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 13 t-test and paired samples correlation. Results showed significant growth and positive correlation in overall writing scores and no significant changes in writer self-perception scores. Overall this study was quite well done. However, as this study did not have a control group, it is difficult to ascertain the full impact of the intervention, apart from the effects of maturation. Cullen and colleagues (2009) used a modified multiple baseline design to investigate the effectiveness of assistive technology (specifically a talking word processor with spell-checker, and word prediction software) on the writing performance of eleven fifth graders with mild disabilities. Students’ writing samples were assessed on mean number of words, mean number of misspellings, accuracy percentage, and total rubric score. The key results showed that the group means increased for all four independent variables. However, due to the small sample size, no statistical analyses were performed. Also, no control group was used, so it is difficult to ascertain the unique impact of the intervention. In another study, Garrett and colleagues (2011) investigated the impact of speech recognition software compared to word processing on the writing of five high school students with physical disabilities. The key results showed that using the software increased the fluency and length of writing, but decreased the accuracy. Using a similar research approach, Mezei and Heller (2012) investigated the impact of word prediction software compared to word processing on the writing of four intermediate students with physical disabilities. The key results showed little to no effect on writing fluency, but it did have some impact on spelling and typographical errors. These studies were also limited by the small sample size, lack of control group and lack of refined statistical analysis. Given the potential to improve writing in certain areas, further research should explore how assistive technology such as general accessibility features on a tablet computer will impact writing when incorporated into other software. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 14 Mobile Devices to Support Learning The development in communication and wireless technologies have resulted in mobile devices (laptops, personal digital assistants, tablet personal computers, cell phones, and e-book readers) becoming widely available, more convenient, and less expensive (Wu et al., 2012). As technology continues to advance, so do the applications and features available on mobile devices, allowing for increased opportunities to use as an educational tool. (Wu et al., 2012). Mobile learning is defined as “learners engaged in educational activities, using technology as a mediating tool for learning via mobile devices accessing data and communicating with others through wireless technology” (Wu et al., 2012, p. 818). Using a longitudinal quasi-experimental methodological approach, Suhr and colleagues (2010) researched the impact of laptops on elementary students’ English Language Arts test scores. They compared changes in English Language Arts scores over a 2 year period on the California Standards Test between the treatment group (54 students with access to their own laptop) and the control group (54 students with no access to a laptop). Using ANOVA and MANOVA analyses, Suhr and colleagues found that students with individual laptops outperformed students who did not have regular access to laptops on changes in their total English Language Arts score. This change score was significant for writing strategies (p < .05) and literary response and analysis (p < .01). Multiple regression analyses revealed that the treatment variable was found to be a significant predictor for changes in the total ELA score, explaining 3% of the variation. Additionally, the treatment variable explained 4% of the change in the literacy response and analysis scores, and 7% of the variation in change in the writing strategies scores, with a small to moderate effect. Overall, this study seemed quite objective and thorough using quite sophisticated statistical measures for analysis. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 15 Wu and colleagues (2012) completed a meta-analysis seeking to determine the purposes, methodologies, and outcomes of mobile learning studies; the main types of devices used in assisted learning; disciplines involved in mobile learning; and highly cited articles in mobile learning. Using rigorous retrieval and analysis procedures, the authors reviewed and synthesized 164 studies from 2003-2010. The researchers discovered most studies used either surveys or experiments to focus on effectiveness of mobile devices or mobile learning design. These topics were also the most regularly cited studies. Results showed that 86% of studies described positive research outcomes and only 4% and 1 % respectively described neutral and negative outcomes. At the time of the study, cellphones and personal digital assistants were found to be most widely used in educational contexts. However, Wu and colleagues speculate that the type of devices most commonly used would change as technology advances. In a more recent meta-analysis and synthesis, Sung and colleagues (2016) studied the effects of integrated mobile devices in teaching and learning through a review of 110 experimental and quasi-experimental journals published from 1993-2013. They sought to provide an overview of the status of the use of mobile devices in educational empirical studies, to quantify the overall effectiveness of integrating mobile devices, to determine the effect of moderator variables that influence the effects of mobile learning, and to synthesize the advantages and disadvantages of mobile learning in levels of moderator variables based on content analysis of articles. Results showed an overall moderate effect of utilizing mobile devices in education. This effect of using mobile devices was better than not using mobile devices, as well as better than using desktop computers. When analyzing the effect of moderator variables, the authors discovered many different combinations of intervention durations. Also, hardware and software have been studied with a variety of ages of users, intervention settings, ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 16 teaching methods, and subject content. In addition, mobile devices were found to be more effective than laptop computers. In terms of teaching strategies used with mobile devices, inquiry-oriented learning was more effective than lectures, self-directed study, cooperative learning, and game-based learning. As well, informal educational settings were more effective than formal settings. Finally, medium and short-term interventions were more effective than long-term interventions. Consequently, a number of variables were identified that impact the intensity of effectiveness of mobile technology intervention studies. This helps to justify the choice of a mobile device instead of a laptop computer, as well as a short-term intervention period in the present research. While these results give some justification for the potential impact of mobile devices on learning, the number of studies included in this meta-analysis that incorporate tablet computers is slight, despite the fact that tablet computers are mentioned by authors as one form of mobile devices. Again, tablet computers were only newly released at the time of this study. Potential of Tablet Computers to support Learning for Children with Disabilities The release of tablet computers, such as the iPad in 2010, has added even more dimensions to the potential impact of technology on learning. At this point, much of the research on the impact of tablet computers on learning is just emerging. In order to explore the effectiveness of tablet computers for children with severe disabilities, Cumming and colleagues (2014) employed an action research project that looked at two main research questions. First, the researchers sought to answer how iPads can assist and enhance learning for students. Second, they questioned the perception of teachers and their students with developmental disabilities in regard to using the iPad as an instructional tool. To begin with, the five participant teachers and research team outlined and participated in a process for professional development pertaining to ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 17 both action research and using the iPad. As a part of the professional development, the teachers also attended biweekly team meetings to share and collaborate. New units of learning were created that incorporated iPad applications chosen for each student based on the students’ Individual Education Plan goals and course content. They then integrated iPads into daily learning for four students with developmental disabilities. The teachers worked individually with each student to teach them how to use each new application. Data were collected through journal type entries on a teacher blog whereby each teacher documented experiences, focus group discussions, and video interviews with both students and teachers. The researchers analyzed the data using inductive content analysis. Elo and Kynga (2008) describe this analysis method as “a systematic research method, which provides objective means to describe phenomena by analyzing content via creating content-related categories” (as cited in Cumming et al., 2014, p. 162). Findings suggest that the iPad improves learning by enhancing instruction, ease of access to the curriculum, providing real-world learning, improving student work and helping students to become more independent learners. Both teachers and students found that iPads were effective and motivating tools for learning. Overall the aforementioned study seems to be very well done. The authors used quite a thick, rich description of setting, participants, procedure and process, as well as analysis of data given. The research design seemed to match their purposes well. The authors were very transparent in identifying limitations of their study, such as the necessity of both qualitative and quantitative methods to more fully evaluate the effectiveness of iPads on student achievement. They also identified that their population was of high SES status and further research is needed across all SES levels as well as more diverse contexts and student populations. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 18 There has been an increased focus on using assistive technology on mobile devices to support the acquisition of literacy skills for students with severe disabilities (Spooner et al., 2015, p. 52). Using a multiple probe across participants design, Spooner and colleagues (2015) used an iPad to examine the use of systematic instruction using a evidence-based shared story format, defined as a method of accessing developmentally appropriate literature through readerlistener interaction as a story is read aloud (Hudson & Test, 2011). The intention of the research was to increase acquisition and generalization of emergent literacy skills and listening comprehension. Five elementary school students with severe cognitive disabilities participated in an intervention program of 44-fifteen minute sessions over approximately five months. The researchers analyzed effectiveness of the intervention based on the difference between baseline and intervention data for each of the participants. The number of correct responses on a task analysis was measured for each session prior to, during, and after an intervention and plotted on a line graph. The graphs were then interpreted. Results showed that all five students were able to demonstrate mastery of the emergent literacy skills taught, and to maintain that mastery in the maintenance phase. Additionally, all five students increased in the number of correct responses to comprehension questions. Furthermore, social validity was measured using a survey, in which eight stakeholders unanimously strongly agreed to all items regarding the valuableness, usefulness, practicality, and generalizability of the intervention. The student participants themselves also responded quite positively to a student Likert-scale survey, indicating that they “loved the iPad2” or “wished they could use the iPad2 all the time at school” (Spooner et al., 2015, p. 40). Although the small sample size in this study does not allow for generalization, beginning evidence is nonetheless given for the potential impact of tablet computers to support ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 19 learning for students with severe disabilities, especially when paired with evidence-based instruction. The Use of Tablet Computers to Support Language Arts In one exploratory study, Hutchison, Beschorner, and Schmidt-Crawford (2012) sought to understand the viability of using iPads to support and enhance literacy instruction. iPads were introduced to a fourth grade classroom of 23 students for a three week period, in which the teacher enhanced the students learning and provided opportunities to learn new literacy skills associated with technology. For each literacy lesson identified, the researchers selected an appropriate iPad application to meet the curriculum outcomes. As a result of this study, the researchers were able to identify many advantages of iPads to enhance literacy instruction. Students were able to navigate the iPad with minimal teacher instruction. As the iPad was found to power on and off quickly and was able to be stored easily, the teacher was able to spontaneously think of ways to integrate the iPad into learning. As well, students were able to work collaboratively to solve problems, which led to increased conversation and enhanced learning. Differentiating assignments, which is providing different avenues of learning to students in order to reach a common learning goal, was easy due to the many different types of applications and the many features available in the applications themselves. Finally, the iPad could be programmed to display in many different languages. The researchers also listed several disadvantages. They found that while manipulation was difficult in some of the applications at times, the touchscreen was sensitive and caused students to engage functions unintentionally at other times. Teachers also had to troubleshoot problems with the technology at times. The teacher also had to adjust to a novel way of receiving and reviewing student work. Finally, some applications did not allow the user to save work, and ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 20 applications for word documents were found to have limited features. Despite these few drawbacks, they came to the conclusion that using the iPads for literacy instruction supported student learning and engagement, and led to creative ways of responding to text while allowing unique features for each user. In the aforementioned exploratory study, not much detail was given regarding the method of data collection or analysis in this study so it is difficult to determine the appropriateness and effectiveness of the research design. More intentional methods of data collection and analysis should have been used or reported on. It is important to note however that the researchers used an exploratory study, likely due to the fact that prior research is almost nonexistent. An exploratory study is necessary in order to help gain more insights and information for further research. Despite the lack of sophisticated method, the aforesaid advantages listed give some evidence for the potential of iPads to enhance literacy learning in students. The study is valuable in this sense as it helps to build a case for the necessity of further research in a very new area. In a more recent study, Bebell and Pedulla (2015) attempted to quantify the impact of 1:1 iPad use, in which every child has access to their own iPad, on student achievement in Language Arts and Math, using two different experimental studies. In the first study, the researchers conducted a 9-week pre/post randomized control trial whereby eight kindergarten classes used literacy and numeracy applications and eight different kindergarten classes used traditional materials. Student achievement was measured through a district assessment battery at the beginning and end of the trial. Results showed stronger literacy gain in the iPad group, but this gain was only statistically significant for one out of the ten subtests: Hearing and Recording Sounds in Words. The second study was longitudinal and explored three years of assessment data in grades kindergarten to grade two, in which students had access to iPads. Effect sizes of ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 21 progress were measured using gain scores and compared to achievement from previous years. The results were ambiguous but did hint at some evidence of increases in Language Arts achievement at the kindergarten level. In writing, greater post-tests gains were seen for the iPad group at the Kindergarten level, but not in grades 1 and 2. Increased pre-post gains were observed for Kindergarten students for Phonemic Awareness during the iPad implementation period as well. There were no gains noted for the iPad group in math achievement. The authors identified the many shortcomings of the methodological design primarily due to the fact that the research was developed around the School Board’s objectives and hence the study was quite limited in scope. These preliminary results hint at potential benefits of iPads to language arts achievement and help to justify further research in the area. The Use of Tablet Computers to Support Writing Using a qualitative narrative inquiry method in which the researchers analyzed “experiences as expressed in lived and told stories of individuals” (Creswell, 2007, p. 54), Sessions and colleagues (2016) investigated the impact of iPad applications on quality of writing. Thirty 5th grade students participated in a nine-week program. While both groups experienced the same instruction on writing strategies, the treatment group used iPad applications to support writing and the control group used only paper and pencil. Data were collected using semistructured interviews between teacher and student, then analyzed and coded. Students’ writing was also collected and coded according to fifth grade core standards. The writing journals of six of the participants were further analyzed as case studies and coded. These six case study students also participated in semi-structured interviews. The key results showed that students using iPad applications wrote stories that were more cohesive, sequential, and included more sensory details than those with paper and pencil. Hence quality of writing was enhanced. Using iPad ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 22 applications also positively impacted motivation to write, allowed students to persist in their writing, and promoted a classroom atmosphere of collaboration. Overall, this study seemed quite rigorously done given the comprehensiveness of the data collection methods, the rich descriptions offered, and the selection of appropriate excepts to support the conclusions drawn. Therefore, this study contributes to the building evidence for the efficacy of iPad applications as a writing intervention. One important area related to writing that has not yet been well researched is the effect of different writing tools (i.e., pencil and paper vs. tablet computers) on writing (Wollscheid, Sjaastad, Tømte, & Løver, 2016). In a pilot study based out of Norway, Wollschied and colleagues began to explore the differences between the two aforementioned writing tools on writing. Using a (2x2) (-1) factorial design (writing instruction * test format), they designed a study to investigate the effect of writing tool and test format (digital vs. pen and paper) on early writing outcomes. A 2x2 design implies four experimental conditions created by the interaction of two variables, each with two levels. In this case, pencil and paper vs. tablet for writing instruction were the two levels of the first variable, and pencil and paper vs. tablet for assessment were the two levels of the second variable. As there was no group that used pencil and paper for instruction and tablet for assessment, only three out of four possible conditions were tested, hence the (-1). Strategic sampling was used to select two primarily schools for their study groups. The first school primarily used pen and paper for literacy tasks and had a sample size of 15. Paper and pencil was used for the writing assessment. The second school comprised of two groups who primarily used tablet computer for writing in class. The first group at this school was assessed using pencil and paper while the second group was assessed using a tablet computer. When ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 23 comparing the group that primarily used paper and pencil with a group that used tablet computer, no significant differences were found when both assessed by paper and pencil. However, when one of the groups that used tablet computer for assessment were compared with the group that used pencil and paper, a significant difference was found. Students assessed by the digital test format wrote faster compared with those assessed with pen and paper, writing approximately 41% more words. No other significant differences were found. Several limitations are present in the study by Wollschied and colleagues (2016), making it difficult to ascertain the true nature of the results. Limited information was given to justify the reasoning for the assessments created. The reliability and validity of the instruments created cannot be ascertained. Furthermore, the methodology was not overly clearly written, making the procedure for the study uncertain. No rationale was given for how the groups were chosen, and all the groups had uneven numbers. Additionally, the independent variables school and condition were confounded and were not controlled by the research design. However, given that this is a pilot study, some evidence for the need for future research is credible. One potential concern with using a tablet computer for writing is the virtual keyboard. Davis and colleagues (2015) highlight several concerns with the virtual keyboards on tablet computers. As compared to the external keyboards of desktop and laptop computers, virtual keyboards lack the conventional resting area of home row and have a smaller size resulting in a more constrained space when selecting keys. These factors may result in slower typing speed. In order to use a virtual keyboard, students must know how to navigate between multiple screens and activate the keyboard when not in use. Due to these factors, Davis and colleagues hypothesized that students using a virtual keyboard would perform worse than both students using a tablet with external keyboard, and students using a laptop on a writing assessment in ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 24 their study. A sample of 387 fifth grade and 439 high school students, randomly assigned to each of the three device and keyboard conditions, responded in essay format to an age-appropriate writing prompt. Data were analyzed with one-way analysis of variance using state test scores as a covariate in order to statistically control for academic proficiency. Contrary to the authors’ hypotheses, data collected revealed no significant difference between each of the three conditions for both grade levels. The authors propose that the length of the essay responses given may have been too short to see an effect. However, it is also possible that the results are valid, suggesting that writing tasks can be assessed on tablet and laptops in a comparable way. Using tablet computers to support writing can be especially effective for students with disabilities. In an action research project, Dunn and colleagues (2016) posited several questions aimed at investigating the effectiveness of iPad applications to improve life readiness skills for nine secondary students with developmental disabilities. One of the areas of life readiness skills analyzed was in the area of literacy (reading and writing) skills. Data were analyzed using quantitative descriptive statistics of Curriculum Based Measurement scores and teachers’ pre and post rating scales. Qualitative data were also collected in the form of collective case studies consisting of semi-structured interviews, bi-weekly journal entries, and observations of student iPad use. The qualitative data were analyzed using a “five-step framework analysis approach” (Dunn et al., 2016, p. 61). The results showed that all students improved in life-readiness skills. Specifically, all students showed gains in their Curriculum Based Measurement scores in reading and writing. Although one could make the argument that maturation could account for gains found in literacy scores, this has less of an impact considering the extremely low level of functioning at the beginning of the project. Given the nature of action research as being ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 25 exploratory, the research design is appropriate and gives credence for further research in this area. In a more recent study, Corkett and Benevides (2016) investigated the impact that iPads have on the writing skills of nine grade six students with a Learning Disability. A visual analysis and paired-sample t-test approach was used to analyze the difference in writing productivity, spelling accuracy, lexical diversity, syntactical complexity, and ideas expressed when using an iPad versus writing by hand. Key results showed that using an iPad significantly improved spelling accuracy and number of ideas expressed. There was an insignificant improvement in writing productivity, number of sentences written, and number of grammatical errors. The authors suggest that there is evidence that iPads have a positive impact on writing and a longterm study may be able to measure this impact more effectively. Summary of Literature Review The extreme paucity of relevant research on assistive technology to support writing is well noted in the literature, despite over 30 years of research that demonstrates some support for the effectiveness of technology to support writing (Batorowicz et al., 2012; Peterson-Karlan, 2011). Yet, studies noted in the above-mentioned reviews do not include current technology, and as such the effects cannot be generalized to technology that has likely advanced or is different than that which was analyzed at the time of the study. Given that assistive technology features on computers have generally been found to have positive impacts for students with special needs (e.g., Cullen et al., 2009; Erdem, 2017; Garrett et al., 2011; Maor et al., 2011; Mezei & Heller, 2012; Perelmutter et al., 2017; Racicot, 2016) it is likely that advances in technology may also lead to improved outcomes, given the past trends (Perelmutter et al., 2017). This is certainly the outcome commonly reported in the small amount of research available on commonly used ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 26 mobile devices such as laptops, cellphones, and tablet computers. Wu and colleagues (2012) examined the impact of mobile devices on learning and found that 86 % of the 164 studies examined reported positive outcomes. Sung and colleagues (2016) found that overall using mobile devices has a greater effect on learning than when using desktop computers, laptop computers, or no mobile device. Suhr and colleagues (2010) noted students with access to a laptop computer for learning in Language Arts outperformed those who did not have access to a laptop. Again, these studies and reviews of studies incorporate very slight information on tablet computers specifically. In the emerging research on tablet computers, some evidence is available to support the use of tablet computers to improve general literacy incomes for students with severe disabilities (Cumming et al., 2014; Spooner et al., 2015) as well as students in general education classrooms (Bebell & Pedulla, 2015; Hutchison et al., 2012). However, all of these studies have shortcomings in their methodological designs. Sessions and colleagues (2016) reported positive outcomes on writing quality with the use of a tablet computer in a general education classroom. Wollscheid and colleagues (2016) found students using a tablet computer outperformed students using pencil and paper on a writing assessment. Again this study suffered from design limitations. Dunn and colleagues (2016) found iPad applications to be effective in improving both reading and writing scores. As this study was exploratory, it also suffered from design limitations as well as a small sample size. Finally Corkett and Benevides (2016) found that using a tablet computer significantly improved spelling accuracy and number of ideas expressed. It is the intention of the present study to contribute to the dearth of experimental evaluative research on extant assistive technology to support the writing of struggling students. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 27 Tablet computers are becoming more commonly used to support writing even though much of the research available does not explore the impact of tablet computers. This study seeks to contribute to the gap in knowledge in this area by using a sophisticated study design and a larger sample size than other comparable studies. Research Problem Given the previous discussion of the importance of discovering effective strategies for struggling writers, as well as the paucity of quality empirical research on the impact of assistive technology on writing for students, the research question being addressed in this study is: Is the iPad application Clicker Docs, in combination with accessibility features, an effective intervention tool to improve writing for struggling writers? This problem will be explored by addressing three specific research questions. 1.) What is the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ writing quality? 2.) What is the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ writing output? 3.) What is the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ attitude towards writing? It is hypothesized that a writing intervention using the iPad application Clicker Docs as assistive technology will have a positive effect on writing quality, writing output, and writing attitude of struggling writers. To achieve this end, struggling writers will be assigned to two groups. In the first treatment phase, Group A will receive the assistive technology intervention while Group B will serve as their control. In the second treatment phase, this will be reversed with Group B receiving the assistive technology intervention while Group A will serve as the ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 28 control. As both groups will be matched in writing ability, it is expected that there will be no significant differences between groups at pre-test. At Post-test 1, it is hypothesized that Group A will show a significant increase in writing quality, writing output, and writing attitude using the iPad as assistive technology. It is hypothesized that Group B will make no significant gains in writing ability, writing output and writing attitude. It is hypothesized that there will be a decrease in writing quality, writing output, and writing attitude between Post-test 1 and Post-test 2 for Group A because they will not again have access to the iPad assistive technology, but that a significant increase will occur for Group B at Post-test 2. The hypothesized outcomes are illustrated in Figure 1. The manner in which the hypotheses will be tested is described in the Writing Scores subsequent chapter. Pre-Test Post-Test 1 Group-A Post-Test 2 Group-B Figure 1. Hypothesized writing assessment outcomes at each testing period. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 29 CHAPTER 3: RESEARCH DESIGN, METHOD, AND PROCEDURES Research Design This study used a switching replications quasi-experimental design. This design is often considered to be one of the strongest of research designs with respect to internal validity (Trochim, 2006). Guthrie and Klauda (2014) describe this design as when: Each student participates in the treatment group and the control group. The implementation of the treatment is repeated or replicated. In the repetition of the treatment, the two groups switch roles; the original control group becomes the treatment group in phase 2, while the original treatment group acts as the control. (p.392) In the first phase of the study, all participants were pre-tested on writing quality, writing output, and attitude towards writing using only paper and pencil. In the second phase, Group A participated in a 6-week assistive technology intervention using the iPad application Clicker Docs and built in tablet accessibility features (the treatment), while Group B used only paper and pencil for writing tasks (the control). At Post-test 1, as Group A had now been trained to use assistive technology, they were allowed the use of the iPad application Clicker Docs and built in tablet accessibility features for Post-test 1, while Group B used paper and pencil. The assessment itself was identical; the only difference was the format that students used to write the assessment. In the third phase, Group B received the same 6-week iPad application Clicker Docs and built in tablet accessibility features intervention (the treatment), while Group A used only paper and pencil for writing tasks (the control). At Post-test 2, both groups were assessed again using the same instruments. This time Group B was allowed the use of the iPad application Clicker Docs and built in tablet accessibility features, while Group A used paper and pencil in order to serve as the control group. See Figure 2 for clarification of this design. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 30 This design is considered to control threats to internal validity and enhance internal and external construct validity, as well as control for individual differences more effectively as students are essentially compared with themselves (Guthrie & Klauda, 2014). This design is often necessary in an educational context, wherein students should not ethically be denied access to a potential intervention. It also allows for less resources as the schools only need to employ enough resources for half of the students at a time (Guthrie & Klauda, 2014). M O0 M O0 X O1 O0 O0 X O1 Legend M –Matched assignment O0 - Assessment with paper and pencil O1 - Assessment using assistive technology X – iPad application Clicker Docs and built in tablet accessibility features intervention Figure 2. Switching Replications Quasi-Experimental Design A matching process was used in order to create treatment and control groups. The participants were assigned to each of the two groups, matched by ability as determined by a combined screener assessment score. Scores were ranked from worst to best for the participants and participants were assigned to each of the two groups with an alternating sequence pattern. Several of the students in the sample had assigned educational assistants. In order to keep the same two educational assistants for both groups in the interventions sessions, it was necessary for two students with similar scores to be switched from one group to the other. One student that was initially in Group A after the alternating sequence pattern was placed in Group B, and another student in Group B was placed in Group A. This trade also allowed the matched groups ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 31 to be closer in mean grade level (see Table 2). No significant difference in writing ability was found between groups according to their pre-screener scores (t (20) = 0.46, p = 0.65). Participants The target population of this research is upper primary and intermediate students who are struggling writers. The population includes students from rural areas in western Canada who attend small independent schools. The target population includes children with Learning Disabilities or other disabilities that result in poorer writing ability. The sample was drawn from a small Christian school in rural BC with 75 students from Kindergarten to grade seven. This particular school had a proportionally high percentage of children with disabilities and special learning needs. For the purposes of this study, intrinsic case study sampling was used. Mertens (2015) defines this sampling procedure as “when a particular case is of specific interest such that the case is in essence already decided before the research begins” (p. 334). The sample size was determined by the technology available. At the time of the study, this context had access to 11 iPads, so each group size could not exceed 11. Therefore, the sample size included two groups of 11, for a total of 22 participants. All of the students in the school from grades 2-7 completed several screening tests to measure overall literacy skills. To determine an appropriate sampling pool, data from two of these assessments were used. The first assessment was designed to measure writing quality. It was developed by another school district in the province and then revised by a group of teachers in the research context. Students were given a picture prompt and a graphic organizer for planning. Students were then given 10 minutes to discuss ideas in small groups and plan for writing. Next, students were given 30 minutes to write a response to the picture prompt. Ten additional minutes were given for editing. Writing was assessed using a rubric that uses a 4-point ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 32 scale in the areas of ideas, organization, sentence fluency, word choice, voice, and conventions. The rubric is slightly different for each grade level as it is based upon grade level standards. The scores for each area were added together to give a total score for overall writing ability. As such the overall score on this assessment could not exceed 24. A student that is considered to be meeting expectations in all areas needs an overall score of at least 18. Several students that were unable to complete this assessment without adult assistance were assigned a score of zero for the purposes of this study. The second screening instrument was the Feifer Assessment of Reading (FAR) screener (Feifer, 2015). This is a short form of a standardized test designed to be a comprehensive evaluation of reading and related processes. The FAR assessment was normed on 1074 students, including students with learning disabilities and intellectual disabilities (Feifer, 2015). The majority of the subtests show median reliability coefficients in the upper .80s and .90s (Feifer, 2015). Given the philosophy that “the skills of reading and writing develop both concurrently and interrelatedly in young children” (Parette et al., 2009, p. 355), and the availability of data for each student, it was considered beneficial to include these data in the screener assessment. The FAR assessments resulted in an overall standard score for every student. To derive a combined screener score for both the school district writing screener and the FAR screener, the writing screener scores were scaled so that both assessments had an identical mean score. The final scores were added together, resulting in the combined screener score as listed in Table 2. When selecting participants for this study, two criteria were followed in order to identify the student as a struggling writer: 1.) lowest combined score on the screening assessments and/or 2.) A diagnosis of Specific Learning Disorder with impairments in written expression. For ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 33 recruitment purposes, students with a diagnosed learning disability in written expression were first invited to participate in the study. Of the remaining students, the 18 identified as having the poorest combined screener score were selected from these grade levels to participate in this study. Consent was not received from three of these potential participants, so the next three participants were invited to participate, as they also demonstrated poor average scores. All of the participants selected received a below average standard score on the FAR screener assessment and performed below grade level on the school district writing screener. Table 1. Gender of Participants. Group A Gender Number of Students Group B Male Female Male Female 6 5 5 6 Table 2. Mean Scores by Group Group Mean Grade FAR Screener 81.18 School District Writing Screener 5.64 Mean Combined Screener Score 130.44 Group A 4.27 Group B 4.18 78.55 6.91 138.93 Out of 22 students recruited, four had diagnosed Learning Disabilities in written expression, one had a diagnosed Learning Disability and a perceptual disability, eight had various low-incidence disabilities, and 12 had individual education plans that included goals pertaining to writing. Prior to the study, none of the participants was using assistive technology ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 34 for writing. The students were from a variety of ethnic backgrounds including Caucasian, First Nations, Polynesian, Haitian, and Hispanic. See Tables 1, 2, and 3 for more detailed participant data. Table 3. Diagnosed Disabilities of Participants. Group A Learning Disability Attention Deficit Hyperactivity Disorder Oppositional Defiant Disorder Developmental Coordination Disorder Mild Intellectual Disability Fetal Alcohol Spectrum Disorder Language Disorder Group B Learning Disability Attention Deficit Hyperactivity Disorder Oppositional Defiant Disorder Developmental Coordination Disorder Mild Intellectual Disability Fetal Alcohol Spectrum Disorder Language Disorder Sensory Processing Disorder Microcephaly Anxiety Disorder NOS Mood Disorder NOS Stereotypic Movement Disorder NOS Attachment Disorder juvenile arthritis Global Developmental Delay Instruments To measure writing quality, a class-wide writing assessment was used at Pre-test, Posttest 1 and Post-test 2. This assessment was the same as the writing quality assessment used in the screener, with a few slight differences. To remove the potential confounding variable caused by the spell check feature on the tablet computer application, the conventions subtest was not considered in either the pre or post-test during the study. The rubric included a 4-point scale in the areas of ideas, organization, sentence fluency, word choice, and voice. Each student’s writing was assessed by the researcher as well as another certified teacher. Any discrepancies were discussed and a final grade was mutually decided upon. The same assessors were used for all pre ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 35 and post-tests. A copy of the instructions for the writing assessment can be found in Appendix B. Additionally, the grade level rubrics can be found in Appendix C. To measure writing attitude, the Writing Attitude Survey (Kear et al., 2000) was used at Pre-test, Post-test 1 and Post-test 2. This survey was designed for elementary students. It uses a 4 point-Likert type scale in which students circle one of four Garfield characters whose emotion most closely matches their own for each question. This survey provides raw scores as well as a percentile rank compared to a norming sample. The survey has an internal reliability of 0.88 using Cronbach’s alpha. Sample items from the Writing Attitude Survey can be viewed in Appendix D. To measure writing output, the number of words written on the writing quality assessment in a 40 minute period was counted by two assessors at each testing interval. Research Materials The intent of this research was to study the effect of assistive technology on struggling writers. Given the desire to use assistive technology that are currently more readily available on tablet computers than other platforms, the tablet computer was chosen as the assistive technology device for this study. At the time of the study, the main software choice of the researcher was not yet available on android or windows-based tablet computers. Given this fact, along with the availability of the technology in the research context, the Apple iPad was chosen as the assistive technology device for this study. It is not the intent of the researcher to suggest that the Apple iPad is superior to other types of tablet computers in supporting writing processes of students. For the purposes of this study, Clicker Docs by Crick Software was the primary software chosen. This is a word processor program designed for all ability levels of writers. Clicker software has been used in thousands of school systems around the world since its release in ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 36 1993, and has been translated into ten languages (Crick Software, 2017). Clicker Docs includes assistive technology features such as word prediction, word banks, spell checker, text-to speech technology, and an adapted keyboard. The feature of ‘word prediction’ works as the user types, by enabling vocabulary to be suggested that comes logically in a sentence, or that is spelled similar to what the student is typing. Clicker Docs also includes ‘word banks’. This allows for tabbed vocabulary suggestions in a large library of topics. Word banks can be created by a teacher or downloaded from an online database. A spell-checker feature is also an option. In addition, ‘clear speech’ text-to speech technology allows a sentence to be read to the child once punctuation is used. This is designed to help children self-correct errors. Clear speech also allows for spell-check selections, and words in word banks to be read aloud. ‘Super keys’ is an adapted keyboard that groups keyboard letters into clusters that enlarge when a cluster is tapped. Screenshots of this application can be viewed in Appendix E. Currently tablet computers contain many built-in accessibility features that are part of the stock operating system. To supplement the application Clicker Docs, the built-in accessibility features ‘speech-to text’, ‘split-screen’, and an ‘artificial digital assistant’ were chosen as part of the treatment. The iPad has a built-in accessibility feature called ‘dictation’ that is accessed by a microphone button on the keyboard. This allows speech-to-text technology in which the user may speak into the tablet computer and have their speech converted into text by the tablet computer. As Clicker Docs has its own keyboard, ‘dictation’ is not available within the Clicker Docs application. As speech-to-text technology was considered beneficial to some students in this study, this feature was accessed through a built in application on the iPad. Students used ‘dictation’ to write sentences on the iPad and then copied and pasted their sentences into Clicker ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 37 Docs. In this situation, the iPad accessibility feature of ‘Split-view’ was used so that students could have both applications open at once and work in both applications at the same time. The final built-in accessibility feature covered was an artificial intelligence digital assistant, available on tablet computers. This is accessed on devices using ‘dictation’, in which the user can dictate questions or commands. Students were taught to use this feature in order to learn how to spell words, as a thesaurus, as a dictionary, to seek information on the internet, and to execute commands such as opening an application. Procedure Prior to this study, the researcher as well as other teachers and Educational Assistants in the school completed training in iPad accessibility features and training in using the application Clicker Docs. The researcher then completed additional training in Clicker Docs that is provided within the Clicker Docs application, and wrote lesson plans for the 12 intervention sessions. These plans are provided in Appendix F. Once approval was obtained from the university’s Research Ethics Board, all 22 students first participated in the pre-test. They gathered in a common classroom and first completed the Writing Attitude Survey together, as the researcher read each question aloud. Next, students completed the Writing Quality assessment. The Researcher followed written instructions when delivering this assessment. Students were shown a stimulus picture on the projector screen as well as given a copy of the photo to explore. Next, they were given 10 minutes to plan for story writing in small groups of three to five children using a planning page provided (see Appendix G). Students were encouraged to think of a story they could make with at least two people, things, or animals that speak. They were encouraged to write ideas for their story down as they ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 38 discussed and to ensure that their story ideas contained a beginning, middle, and end. They were then given lined paper and 30 minutes for writing, followed by 5-10 minutes for editing. The following week, the 11 students in Group A were assigned a tablet computer and then completed a 6 week intervention with the researcher that involved two 25 minute intervention sessions per week, while Group B participated in their regular classroom learning activities using pencil and paper. Due to the high level of student needs, two Educational Assistants were also present during all intervention sessions. For each lesson, every student had access to their own tablet computer. The researcher demonstrated on a personal tablet computer that was mirrored on a projector screen so that the students could follow along. During the first few sessions, children became familiar with accessibility features and had opportunities to practice those features in short writing assignments. The next sessions involved learning how to use features of the application Clicker Docs and again practicing in short writing assignments. The remaining weeks involved planning for writing and practicing a variety of short writing assignments using accessibility features as needed in Clicker Docs. The final weeks involved practicing revision of work on the Clicker Docs application. During the 6 week intervention phase, the intervention group was also encouraged to use their tablet computer for all in-class writing tasks. In class practice time for each student was recorded daily and the total was calculated per student at the end of the 6 week intervention. Whenever a student missed intervention sessions due to absences, the researcher met with the student individually to deliver any content missed, prior to the next intervention session. At the end of the 6 week phase, all students completed Post-test 1, with the intervention group using the iPad application Clicker Docs and built in tablet accessibility features for their writing sample, and the control group using only pencil and paper. The post-test was completed ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 39 several days after the intervention session ended. After the post-test, the tablet computers were then reassigned to the 11 students in Group B. For the next 6 weeks the groups switched and the procedure was repeated. As Group A was now the control group, tablet computers were no longer used during in-class writing for Group A and tablet computers were reassigned to Group B. At the end of the project both groups completed Post-test 2. This time Group B used the iPad application Clicker Docs and built in tablet accessibility features for their writing sample, while Group A used paper and pencil in order to serve as the control group. Post-test 2 was also completed within a few days of the intervention ending. Data Analysis A mixed 2x2 repeated measures analysis of covariance (ANCOVA) with pre-test scores as covariate was used to measure each of the three dependent variables. This statistical procedure is described by Howell as being robust and effective for small sample sizes (as cited in Curcic & Johnstone, 2016, p. 73). The within-subject factor was time (Post-test 1 and Post-test 2) and the between-subject factor was group (A and B). The pre-test was used as covariate in order to factor out the effect of the pre-test scores. Dependent variables included writing quality, writing output, and attitude towards writing. Effect sizes were calculated using partial eta squared for each dependent variable. A large effect size is determined to be greater than .35, a medium effect size is greater than .15, and a small effect size is greater than .02 (Cohen, 1992). For each significant interaction, ANCOVAs were used for post-hoc analyses to determine the level of significance between groups for each variable measured. Effect sizes for post hoc ANCOVAs were also calculated using partial eta squared. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 40 Ethics Approval and consent. Research Ethics Board approval was received prior to starting research. A letter of consent was obtained from the Principal of the research context. A letter of consent was obtained for each of the students participating in the study. The project was explained in detail to the parents of participants in the consent letter. Parents were assured that the participation of their child was voluntary and consent could be given or denied based upon what they were comfortable with. Also, they were assured that consent could be revoked at any time. Parents of potential student participants were invited to ask any questions, or discuss any concerns with the researcher. Once consent was obtained, students had the project explained to them in ageappropriate language. Every effort was made to ensure students were in no way harmed. It was explained to students that if they felt distressed in any way, they were welcome to take a break or not participate. Anonymity and confidentiality. Participants were given an alpha-numeric code that corresponds to their identity to preserve privacy. Only the researcher and research assistants had access to this coding system. Computer data files were password protected. Paper copies of data were kept in a locked filing cabinet. Potential conflict of interest. The researcher was a teacher and the Special Education Coordinator at the school where the research took place. She had a pre-existing relationship with all of the participants. The researcher was also a colleague of the research assistants. This was necessary given the unique opportunity for research currently taking place in this context that would otherwise not have been available to the researcher. The perceived success or failure of the intervention may have ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 41 indirectly impacted the perceived effectiveness of the researcher in her job assignment. Also, given the small community, the researcher was a parent of four children in the school. In order to avoid the impact of inhibited free consent due to the power relationship that existed, these four children were not included in the study. Scientific Quality: Validity and Reliability The researcher made every effort to ensure a study that was as reliable and valid as possible given the financial and time constraints. As random assignment was not possible in this research context, a quasi-experimental design was chosen. However, every effort was made to choose a research design that allowed for a control group, thus improving internal and external construct validity, as previously mentioned. To reduce effect of differential instruction, the two groups were matched to ensure there were no significant differences in writing ability. Every effort was made to ensure that the mean age of the two groups were also similar. To minimize effects of history and to improve treatment fidelity, a script was written for intervention sessions to ensure that the intervention was as similar for both groups as possible. To control for instrumentation, the same assessments were used at each testing interval, with the exception of a different picture prompt. Having a different picture prompt helped to reduce the threat of testing or students learning from the previous assessment. However, each picture prompt was similar in style as it was completed by the same artist. The pictures were black and white, and depicted a mysterious situation. To control for experimental treatment diffusion, tablet computers were assigned to the treatment group and were not used by the control group for writing during each phase. Children in the study who were seen to benefit from the intervention will have access to this technology in their future education, and will know how to use this assistive technology for writing, thus enhancing ecological validity. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 42 Efforts were also made to use instruments with high reliability. Reliability for the Writing Attitude Survey was noted earlier. As no standardized writing measure could be used, two assessors were used for the writing quality assessments, in an attempt to improve reliability. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 43 CHAPTER 4: RESULTS The research problem being addressed in this study is whether the iPad application Clicker Docs is an effective intervention tool to improve writing for struggling writers. This problem was addressed by looking at three specific research questions. Question 1: Writing Quality The first research question looked at the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ writing quality. It was hypothesized that after controlling for pre-test, the treatment group would show a significant increase in writing ability as a result of using the iPad application Clicker Docs and accessibility features. It was expected that the control group would not show a significant increase in writing ability. Therefore, after controlling for pre-test, it was expected that the first treatment group (Group A) would perform significantly better in writing quality scores than the control group (Group B) at Post-test 1; and that the second treatment group (Group B) would perform significantly better in writing quality scores than the control group (Group A) at Post-test 2. Therefore, a significant interaction would support this hypothesis. A mixed 2x2 repeated measures analysis of covariance (ANCOVA) with pre-test scores as covariate was used. The interaction effect tested the interaction between the independent variables of group (Group A vs. Group B) and time (Post-Test 1 and Post-Test 2) on the dependent variable of writing quality. First the assumptions were tested. The assumption of homogeneity of variance was met for both Post-test 1 and Post-test 2. Furthermore, no significant differences existed between groups on the pre-test. The assumption data can be viewed in Table 4. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 44 Table 4. Assumptions Tested for Each Writing Assessment Dependent Variable. Homogeneity of Variance F Significance (1,20) Writing Quality Post-test 1 0.29 0.59 Writing Quality Post-test 2 0.52 0.48 Writing Output Post-test 1 4.34 0.05 Writing Output Post-test 2 2.03 0.17 Attitude towards Writing Post-test 1 0.01 0.94 Attitude Towards Writing Post-test 2 1.05 0.32 Differences Between Groups on Pre-test F Significance (1,20) Writing Quality Pre-test 0.06 0.80 Writing Output Pre-test 0.002 0.97 Attitude towards Writing Pre-test 0.03 0.87 Results of the 2x2 ANCOVA showed a very large significant interaction effect of the iPad application Clicker Docs and accessibility features on writing quality, after controlling for pre-test. The independent variables time and group interacted on the dependent variable writing quality. These specific data can be viewed in Table 5. This result is also displayed graphically in Figure 3. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 45 Table 5. Interaction Effects of the Independent Variables on Each Dependent Variable. Main Interaction Effects F Significance Effect Size (Partial ɳ2 ) (1,20) Writing Quality 16.00 0.001 0.46 Writing Output 4.44 0.05 0.19 Attitude towards Writing 0.48 0.50 0.03 Post Hoc Analyses F Significance Effect Size (Partial ɳ2 ) (1,19) Writing Quality Post-test 1 2.58 0.125 0.12 Writing Quality Post-test 2 13.55 0.002 0.42 Writing Output Post-test 1 3.59 0.07 0.16 Writing Output Post-test 2 1.81 0.19 0.09 The first post hoc ANCOVA analyses were completed for Post-test 1, using group as an independent variable, pre-test scores as a covariate, and writing quality composite scores as a dependent variable. The analyses revealed a nonsignificant effect of the iPad application clicker docs and accessibility features on writing quality, after controlling for pre-test at Post-test 1. Even though the result was not significant, perhaps due to small sample size, the treatment had a small to medium effect size on writing quality. The next post hoc ANCOVA analyses were completed for Post-test 2, using group as an independent variable, pre-test scores as a covariate, and writing quality composite scores as a dependent variable. The analyses revealed a large significant effect of the iPad application Clicker Docs and accessibility features on writing quality, after controlling for pre-test at Post- ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 46 test 2. On average, those in the iPad intervention group demonstrated better writing quality than those in the control group, when factoring in pre-test scores (see Table 6). This difference was significant at Post-test 2, but not at Post-test 1. See Table 5 for detailed data. Figure 3. Mean Writing Quality Composite Scores by Group. Question 2: Writing Output The second research question looked at the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ writing output. It was hypothesized that after controlling for pre-test, the treatment group would show a significant increase in writing output as a result of using the iPad application Clicker Docs and accessibility features. It was expected that the control group would not show a significant increase in writing ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 47 output. Therefore, after controlling for pre-test, it was expected that the first treatment group (Group A) would perform significantly better in writing output scores than the control group (Group B) at Post-test 1; and that the second treatment group (Group B) would perform significantly better in writing output scores than the control group (Group A) at Post-test 2. This would be confirmed by a significant interaction effect. Table 6. Mean Scores on Writing Assessment Post-tests. Group A Group B Writing Assessment Significance Effect Size M SE M SE Writing Quality Post-test 1 8.32a 0.52 7.13a 0.52 0.125 0.12 Writing Quality Post-test 2 7.19a 0.31 8.81a 0.31 0.002 0.42 Writing Output Post-test 1 97.39a 25.64 166.07a 25.64 0.07 0.16 Writing Output Post-test 2 147.09a 17.05 114.64a 17.05 0.19 0.09 Attitude towards Writing Post-test 1 61.55a 3.25 59.99a 3.25 >.05 n/a 58.77a 3.44 61.23a 3.444 >.05 n/a (Partial ɳ2 ) *higher score is greater attitude Attitude Towards Writing Post-test 2 *higher score is greater attitude a mean scores are adjusted for covariate A mixed 2x2 repeated measures analysis of covariance (ANCOVA) with pre-test scores as covariate was used. The interaction effect tested the interaction between the independent variables of group (Group A vs. Group B) and time (Post-Test 1 and Post-Test 2) on the dependent variable of writing output. First the assumptions were tested. The assumption of homogeneity of variance was not met for Post-test 1 but was met for Post-test 2. Furthermore, no significant differences existed between groups on the pre-test. The assumption data can be viewed in Table 4. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 48 Results of the 2x2 ANCOVA showed a medium significant interaction effect of the iPad application Clicker Docs and accessibility features on writing output, after controlling for pretest. This specific data can be viewed in Table 5. This result is also displayed graphically in Figure 4. Figure 4. Mean Writing Output Composite Scores by Group. The first post hoc ANCOVA analyses were completed for Post-test 1, using group as an independent variable, pre-test scores as a covariate, and writing output scores as a dependent variable. The analyses revealed a marginally nonsignificant effect of the iPad application Clicker ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 49 Docs and accessibility features on writing output, after controlling for pre-test at Post-test 1. Even though the result was not significant, perhaps due to small sample size, the treatment had a medium effect size on writing output. Contrary to researcher hypotheses, on average, those in the iPad intervention group wrote less overall than those in the control group, when factoring in pretest scores (see Table 6). The next post hoc ANCOVA analyses were completed for Post-test 2, using group as an independent variable, pre-test scores as a covariate, and writing output scores as a dependent variable. The analyses revealed a nonsignificant effect of the iPad application Clicker Docs and accessibility features on writing output, after controlling for pre-test at Post-test 2. This difference was marginally significant at Post-test 1, but not at Post-test 2. See Table 5 for detailed data. Question 3: Attitude towards Writing The third research question looked at the effect of using the iPad application Clicker Docs, in combination with accessibility features, on struggling students’ attitude towards writing. It was hypothesized that after controlling for pre-test, the treatment group would show a significant increase in attitude towards writing as a result of using the iPad application Clicker Docs and accessibility features. It was expected that the control group would not show a significant increase in attitude towards writing. Therefore, after controlling for pre-test, it was expected that the first treatment group (Group A) would perform significantly better in attitude towards writing scores than the control group (Group B) at Post-test 1; and that the second treatment group (Group B) would perform significantly better in attitude towards writing scores than the control group (Group A) at Post-test 2. This would be demonstrated by a significant interaction effect. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 50 A mixed 2x2 repeated measures analysis of covariance (ANCOVA) with pre-test scores as covariate was used. The interaction effect tested the interaction between the independent variables of group (Group A vs. Group B) and time (Post-Test 1 and Post-Test 2) on the dependent variable of attitude towards writing. First the assumptions were tested. The assumption of homogeneity of variance was met for both Post-test 1 and Post-test 2. Furthermore, no significant differences existed between groups on the pre-test. Assumption data can be viewed in Table 4. Results of the 2x2 ANCOVA showed no significant interaction effect of the iPad application Clicker Docs and accessibility features on attitude towards writing, after controlling for pre-test. See Table 5 for specific data. Therefore, we can conclude that the iPad application Clicker Docs in combination with accessibility features does not have an effect on struggling writers’ attitude towards writing. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 51 CHAPTER 5: DISCUSSION The purpose of this study was to experimentally examine the effectiveness of tablet computers as an intervention for improving struggling students’ writing output, writing quality, and attitude towards writing. This research focused on Clicker Docs, an application for iPad, combined with built-in tablet accessibility features as an intervention. Question 1: Writing Quality The objective of the first research question was to determine the effect of assistive technology on writing quality among students with writing difficulties. After controlling for pretest, students in the iPad intervention group demonstrated better average writing quality scores than those in the control group. Although this result was not statistically significant at Post-test 1, it was significant at Post-test 2 with a large effect size. This discrepancy in results can be explained by problems encountered during the administration of Post-test 1. Unfortunately during the first post-test, the Researcher had to leave unexpectedly during the writing sample portion for an urgent medical appointment. After she left, several behavioral incidents occurred. Due to circumstances beyond the researcher’s control, two students in Group B became emotionally distraught and were defiant and oppositional while completing the writing sample portion of the assessment. Given the nature of the diagnoses of these two students, behavioral incidents such as these are common. Another student in group A was highly distracted by these incidents and initially had difficulty completing his writing assessment. While all of the students did eventually complete their writing samples, their emotional state and ability to focus may have affected the quality of their responses. During the administration of Post-test 2, there were no behavioral incidents and all of the students seemed focused on the writing task. All of the students appeared to put forth a good effort. Thus, it is the researcher’s belief that the significant ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 52 results obtained in Post-test 2 reflect the actual effect of the iPad application Clicker Docs on writing quality. These findings align with the results of Racicot (2016), in which a significant growth in overall writing scores was found, following an intervention using a Clicker software application. The researcher’s general observations of the children as they used the iPad application Clicker Docs for their writing samples also confirmed a higher quality of writing. There especially appeared to be an impact on vocabulary. Many of the students appeared to choose their words more carefully, often considering words supplied in the word banks. At times students were able to attempt more complex words with the spelling support offered by the word prediction feature. Several students chose to utilize more complex vocabulary by using the dictation and spell-check features as well. This observation is confirmed by Perelmutter and colleagues (2017), as they also found speech-to-text interventions to have positive outcomes, and spell-check features to have large beneficial effects. There was also a positive impact on the quality of ideas expressed when using the iPad application Clicker Docs. Quite a few of the weaker students who struggled with writing output using a pencil were able to get more ideas on paper when using features of the iPad application Clicker Docs. This observation is supported by the findings of Batorowicz and colleagues (2012), in which technology was found to increase independence in writing, and to have more of an impact for low performers. Corkett and Benevides (2016) also noted an improvement in number of ideas expressed when an iPad was used. Finally, it was the general impression of the researcher and research assistant that students using the iPad application Clicker Docs showed superior sentence structure and quality. Students in the intervention group were able to write less run-on sentences than students using paper and pencil. Overall, sentences on Clicker Docs were ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 53 more concise and coherent than many of the sentences fabricated with paper and pencil. This may largely be due to the text-to-speech feature on Clicker Docs in which sentences are read back to the user as they are completed, or when prompted by the user. This observation is supported by the findings of Perelmutter and colleagues (2017), who noted a small, positive effect of text-to-speech interventions. Moreover, the observation of improved sentence quality is supported by the results obtained by Sessions and colleagues (2016) who found that students using iPad applications wrote stories that were more cohesive and sequential than those using paper and pencil. Question 2: Writing Output The objective of the second research question was to determine the effect of assistive technology on writing output among students with writing difficulties. Some research has shown that using technology to support writing leads to increased writing output (Cullen et al., 2009; Garrett et al., 2011; Wollscheid et al., 2016). Conversely, the opposite effect was found in the present study, though with marginally significant results likely due to the small sample size. Contrary to researcher hypotheses, those in the iPad intervention group wrote less overall than those in the control group, when factoring in pre-test scores. This difference was marginally significant at Post-test 1 and not significant at Post-test 2. Here again, some of the discrepancy in results between Post-test 1 and Post-test 2 can likely be attributed to the problems in the administration of Post-test 1. Students who were involved in or distracted by behavioral incidents were not engaged in as much time writing as they likely would have in a more calm and regulated emotional state, as was the case in Post-test 2. One factor that may have accounted for the opposite hypothesized writing output result is the typing experience and ability of the research participants. Generally the students in the study ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 54 were more comfortable and experienced with using a paper and pencil for writing than typing. Some of the youngest students had minimal experience with typing or had never been taught to type properly. Many of the older students were quite slow and awkward with typing. Even though students had several options on the software to support writing output, the students generally also had less experience in using these areas. For example, many students using speech-to-text, or the adapted virtual keyboard options had never before used these features for writing. Perhaps if the participants had more experience using technology for writing prior to the study, the writing output outcome may have been different. It is logical to assume that students may become more proficient in increasing their writing output using assistive technology as they become more practiced with using the technology. In addition, word count may also have been affected by “lagging” devices as a result of technical problems with our school network. After the study was completed it was discovered that our school network was improperly configured for multiple devices, which resulted in interference between devices at times and caused them to work more slowly. While this did not seem to be a major factor, it is difficult to ascertain the full impact that this may have had on writing output. Another factor that may have impacted writing output was the extra time needed to implement the features of the iPad application Clicker Docs. If students chose to use features of the program designed to support writing, it often took time to do so. For instance, deciding to use a more sophisticated word by looking for synonyms in the Word bank does require more time than just writing a sentence with whatever word comes to mind. Finally, given the overall greater quality of writing samples generated by the iPad intervention group, it is important to note that less words does not always mean less quality. Even though the writing samples on the iPad were substantially shorter, they were often more ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 55 concise. The text-to-speech feature, in which students’ writing was read back to them, may have helped to identify run-on sentences and to trim back unnecessary details. It may also be that if using technology took more effort to write due to less practice using it, students may have thought more carefully about what they were going to write before they wrote it, instead of writing as they thought of ideas. In contrast, many of the paper and pencil writing samples contained rambling sentences that lacked clear direction, purpose, and organization. Question 3: Attitude towards Writing The objective of the third research question was to determine the effect of assistive technology on attitude towards writing among students with writing difficulties. Much research has shown that using technology to support writing leads to increased motivation, engagement, and attitude towards writing (Batorowicz et al., 2012; Cumming et al., 2014; Hutchison et al., 2012; Sessions et al., 2016; Spooner et al., 2015). However, no effect was found in the present study. Contrary to the researcher’s hypothesis, there was no significant interaction effect of the iPad application Clicker Docs and accessibility features on attitude towards writing after controlling for pre-test. While the results of this study would appear to indicate that the iPad application Clicker Docs in combination with accessibility features does not have an effect on struggling writers’ attitude towards writing, it is the researcher’s considered opinion that these results may not hold much validity. Several factors may have accounted for the results obtained in the present study. Many of the factors are directly related to the instrument used to measure the variable, the Writing Attitude Survey (Kear et al., 2000). It is the researcher’s impression that the participants demonstrated testing fatigue due to multiple administrations of the same instrument in a relatively short period of time. With each subsequent administration of this survey, the students ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 56 showed increasingly negative attitudes during completion. Comments such as “ugh, why do we have to do this again, we did this already” were observed by the researcher. The fact that both groups’ average attitudinal scores showed a slight downward trend from pre-test supports this hypothesis. This survey may not have been designed for multiple uses over a short period of time. Additionally, the Writing Attitude Survey may not have adequate construct validity for this specific research. The responses students gave on the survey did not match general observations of the researcher and school staff during the course of the present study. Students were observed to be excited and motivated while using the iPads for writing tasks and eager to begin writing when iPads were being used, both in the classroom context as well as during intervention sessions. Lastly, this survey was designed for writing in general and not necessarily for writing with technology. Consequently, it is difficult to determine how much participants were able to generalize their attitudes towards writing during the course of this study to their responses on the survey. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 57 CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS Summary The present study sought to determine if the iPad application Clicker Docs is an effective intervention tool to improve writing for struggling writers by looking at three specific components of writing: writing quality, writing output, and attitude towards writing. As hypothesized, using tablet computers as assistive technology resulted in higher quality writing than when tablet computers were not used. It is the conclusion of the researcher that the iPad application Clicker Docs in combination with accessibility features, is an effective assistive technology tool for struggling writers. Writing quality scores showed significant improvement when using Clicker Docs and accessibility features as assistive technology. Although the writing samples were shorter, they were overall much better written. Using less words, students were often able to express more ideas of higher quality in a concise manner. As conventions were not considered, this increase in writing quality had little to do with spelling, grammar, and punctuation as may be commonly expected. No significant interaction effect of the iPad application Clicker Docs and accessibility features was found for attitude towards writing. These results do not align with either the researcher’s observations during the intervention and post-test periods, or with prior research. This highlights a need for further research on attitude towards writing before conclusions can be fully drawn. Implications of Research Several past studies have highlighted the necessity of evaluating current improvements in up to date technology (Batorowicz et al., 2012; Perelmutter et al., 2017; Peterson-Karlan, 2011;). As Clicker Docs for iPad was only released in 2014, and versions are currently being developed ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 58 for other devices, this study contributes greatly to research by providing insight into the effectiveness of a fairly new application. This research also contributes to alleviating the dearth of research on tablet computers by building upon the findings of pilot and exploratory studies. The present study found a significant effect of the iPad application Clicker Docs on writing quality under controlled assessment conditions. A large effect size was found, demonstrating the practical significance and usefulness of such an intervention in a school context. There is much knowledge gained from this study that could be applied to educational practice. Perelmutter and colleagues (2017) highlight the need for technology interventions to be customized to each individual child in order to be deemed effective. In the present study, the participants were able to select from a number of different features available in the iPad and the application Clicker Docs. Children were able to choose the supports best suited for their needs. For example, a child with fine motor concerns could choose to use the adapted keyboard or speech-to-text technology. A child with a lower reading level could select a smaller number of words to be displayed in the word prediction screen, whereas a student with a higher reading ability could select a higher number of words to be displayed. This has great practical value for using the technology among a wide variety of learners in classroom while still meeting each learner’s individual needs. One of the barriers to successfully using assistive technology to support learners in a classroom is lack of teacher training in how to use the technology (Flanagan et al., 2013). The major emphasis on the intervention portion of the current research was teaching students to be comfortable with using the technology. As all of the intervention lessons are included in Appendix F, it would be fairly easy for teachers in other settings to use the lesson plans in order to train students or staff to utilize this specific technology. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 59 Limitations There are some noteworthy limitations of the present study. For the purposes of this research it was not possible to utilize random assignment for control groups, or to randomly select a research context. Even though the sample size was larger than other comparable studies, the small sample size was a limitation. Financial and time constraints would not allow for a fullscale study. Twelve weeks was the maximum amount of time available for the study in the identified school context. However, a six week intervention period may not have been enough time for participants to receive the maximum impact of the technology intervention. Given the availability of a school project in an accessible context, a unique opportunity for research was presented and helps to justify these limitations of research design. Despite the efforts to use sophisticated methods of data collection and analyses, there were several limitations to the instrumentation used. This was most notable for the variable of attitude towards writing. As previously mentioned, the survey used for this did not adequately measure the intended variable of attitude towards writing. The instrument used to measure writing quality could also have been improved upon. It may have been preferable to utilize a standardized assessment for measuring writing quality to reduce the effect of evaluator subjectivity. However, most standardized writing assessments require individualized administration and scoring for each testing interval. Time constraints would not allow for this extent of assessment. In addition, the study was limited by utilizing only quantitative data. Cumming and colleagues (2014) highlight the necessity of using both qualitative and quantitative methods to fully evaluate the effectiveness of iPads on writing achievement. It would have been of benefit to include qualitative data if time and finances had allowed for a mixed methods approach. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 60 Keeping a diary of the researcher’s informal observations as well as conducting surveys or interviews of teachers, parents, and students to gain insight into the perceptions of the project would have been extremely valuable to this research. Thus, a mixed methods approach would have perhaps given a more complete picture of the actual effects of the intervention, especially when considering attitudes towards writing. Although every attempt was made to control for extraneous variables in the research design, there were a few factors that may have confounded the results. Due to the complex nature of student diagnoses and needs, negative behavior experienced during Post-test 1 may have been a confounding variable. Network lagging during the writing assessments may have impacted the results. This, however is a realistic problem out of study context as well, and should be considered a drawback wherever technology is used. This thus adds to the ecological validity of the study’s findings. Considerations for Future Research There are a number of ways to improve and build upon the present research. In order to improve methodology, future studies would benefit from considering a mixed method approach to data collection. When considering writing output, it would also be beneficial to measure a student’s proficiency and experience with typing and tablet use prior to beginning the study. It would improve the current research to increase the sample size, and consider intervention groups across several different school settings. It may also be interesting to compare multiple age groups to see if there is a difference in writing output. To minimize the confounding effect of behavior, it may be pertinent to exclude children with emotional and behavioral diagnoses from future research samples. Finally, in order to ensure technology is working at maximum capacity, ensure that the network in the research context is properly configured for multiple device use. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 61 Concluding Thoughts Assistive technology has been found to have a positive effect on writing quality in numerous other studies (Corkett & Benevides, 2016; Maor et al., 2011; Perelmutter et al., 2017; Suhr et al., 2010; Sung et al., 2016; Wu et al., 2012). The findings of the present study give further evidence to this conclusion. Throughout the research period, the researcher was encouraged by the results observed. Several months after the study was complete, many of the research participants are continuing to use the iPad application Clicker Docs and accessibility features in the classroom, and continue to grow in their writing ability. Plans have been made to run a second intervention program for some of younger students, and newly enrolled students who did not get to participate. There are also a number of staff members in this context who would like additional training in using the application. The iPad application Clicker Docs has tremendous practical value in schools and great potential as assistive technology for struggling writers. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 62 REFERENCES Adebisi, R. O., Liman, N. A., & Longpoe, P. K. (2015). Using assistive technology in teaching children with learning disabilities in the 21st century. Journal of Education and Practice, 6(24), 14-20. Akpan, J. P., & Beard, L. A. (2013). Overview of assistive technology possibilities for teachers to enhance academic outcomes of all students. Universal Journal of Educational Research, 1(2), 113-118. Alnahdi, G. (2014). Assistive technology in special education and the universal design for learning. Turkish Online Journal of Educational Technology - TOJET, 13(2), 18-23. Asaro-Saddler, K., Knox, H. M., Meredith, H., & Akhmedjanova, D. (2015). Using technology to support students with Autism Spectrum Disorders in the writing process: A Pilot Study. Insights on Learning Disabilities, 12(2), 103-119. Batorowicz, B., Missiuna, C. A., & Pollock, N. A. (2012). Technology supporting written productivity in children with learning disabilities: A critical review. Canadian Journal of Occupational Therapy, 79(4), 211-224. Bebell, D., & Pedulla, J. (2015). A quantitative investigation into the impacts of 1:1 iPads on early learners' ELA and math achievement. Journal of Information Technology Education: Innovations in Practice, 14, 191-215. Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159. doi:10.1037/00332909.112.1.155 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 63 Corkett, J. K., & Benevides, T. (2016). IPad versus handwriting: Pilot study exploring the writing abilities of students with learning disabilities. Journal of International Special Needs Education, 19(1), 15-24. Creswell, J. W. (2007). Qualitative inquiry and research design: Choosing among five approaches (2nd ed.). Thousand Oaks, CA: Sage. Crick software (2009). About Crick software. Retrieved April 23, 2017, from http://www.cricksoft.com/us/about/index.aspx Cullen, J., Richards, B., & Frank, C. L. (2009). Using software to enhance the writing skills of students with special needs. Journal of Special Education Technology, 23(2), 33-44. Cumming, T. M., Strnadová, I., & Singh, S. (2014). IPads as instructional tools to enhance learning opportunities for students with developmental disabilities: An action research project. Action Research, 12(2), 151-176. doi:10.1177/1476750314525480 Curcic, S., & Johnstone, R. S. (2016). The effects of an intervention in writing with digital interactive books. Computers in the Schools, 33(2), 71-88. Davis, L. L., Orr, A., Kong, X., & Lin, C. (2015). Assessing student writing on tablets. Educational Assessment, 20(3), 180-198. doi:10.1080/10627197.2015.1061426 Delacruz, S. (2014). Using Nearpod in elementary guided reading groups. TechTrends: Linking Research & Practice to Improve Learning, 58(5), 62-69. doi:10.1007/s11528-014-0787-9 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 64 Dunn, M. (2015). Struggling writers' use of iPad art and text apps for story writing. Canadian Journal of Action Research, 16(2), 6-21. Dunn, M., Barrio, B., & Hsiao, Y. (2016). Do iPad applications help students with developmental disabilities improve life-readiness skills? Canadian Journal of Action Research, 17(1), 51-75. Erdem, R. (2017). Students with special educational needs and assistive technologies: A literature review. Turkish Online Journal of Educational Technology - TOJET, 16(1), 128-146. Feifer, S.G. (2015). Far screen: Feiffer assessment of reading screener form. Lutz, Fl: PAR. Flanagan, S., Bouck, E. C., & Richardson, J. (2013). Middle school special education teachers' perceptions and use of assistive technology in literacy instruction. Assistive Technology, 25(1), 24-30. Garrett, J. T., Heller, K. W., Fowler, L. P., Alberto, P. A., Fredrick, L. D., & O'Rourke, C. M. (2011). Using speech recognition software to increase writing fluency for individuals with physical disabilities. Journal of Special Education Technology, 26(1), 25-41. Guthrie, J. T., & Klauda, S. L. (2014). Effects of classroom practices on reading comprehension, engagement, and motivations for adolescents. Reading Research Quarterly, 49(4), 387416. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 65 Hudson, M. E., & Test, D. W. (2011). Evaluating the evidence base of shared story reading to promote literacy for students with extensive support needs. Research and Practice for Persons with Severe Disabilities, 36(1-2), 34-45. Hutchison, A., Beschorner, B., & Schmidt-Crawford, D. (2012). Exploring the use of the iPad for literacy learning. Reading Teacher, 66(1), 15-23. Johnson, G. M. (2013). Using tablet computers with elementary school students with special needs: The practices and perceptions of special education teachers and teacher assistants. Canadian Journal of Learning and Technology, 39(4), 1-12. Kagohara, D. M., van der Meer, L., Ramdoss, S., O'Reilly, M. F., Lancioni, G. E., Davis, T. N., & ... Sigafoos, J. (2013). Using iPods and iPads in teaching programs for individuals with developmental disabilities: A systematic review. Research in Developmental Disabilities: A Multidisciplinary Journal, 34(1), 147-156. Kear, D. J., Coffman, G. A., McKenna, M. C., & Ambrosio, A. L. (2000). Measuring attitude toward writing: A new tool for teachers. Reading Teacher, 54(1), 10. Larabee, K. M., Burns, M. K., & McComas, J. J. (2014). Effects of an iPad-supported phonics intervention on decoding performance and time on-task. Journal of Behavioral Education, 23(4), 449-469. doi:10.1007/s10864-014-9214-8 Maich, K., & Hall, C. (2016). Implementing iPads in the inclusive classroom setting. Intervention in School and Clinic, 51(3), 145-150. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 66 Maor, D., Currie, J., & Drewry, R. (2011). The effectiveness of assistive technologies for children with special needs: A review of research-based studies. European Journal of Special Needs Education, 26(3), 283-298. Mertens, D. M. (2015). Research and evaluation in education and psychology (4th ed.). Thousand Oaks, CA: Sage. Mezei, P. J., & Heller, K. W. (2012). Effects of word prediction on writing fluency for students with physical disabilities. Physical Disabilities: Education and Related Services, 31(1), 3-26. Musti-Rao, S., Lo, Y., & Plati, E. (2015). Using an iPad app to improve sight word reading fluency for at-risk first graders. Remedial and Special Education, 36(3), 154-166. doi:10.1177/0741932514541485 Parette, H. P., Hourcade, J. J., Dinelli, J. M., & Boeckmann, N. M. (2009). Using "Clicker 5" to enhance emergent literacy in young learners. Early Childhood Education Journal, 36(4), 355-363. Perelmutter, B., McGregor, K. K., & Gordon, K. R. (2017). Assistive technology interventions for adolescents and adults with learning disabilities: An evidence-based systematic review and meta-analysis. Computers & Education, 114,139-163. Peterson-Karlan, G. R. (2011). Technology to support writing by students with learning and academic disabilities: Recent research trends and findings. Assistive Technology Outcomes and Benefits, 7(1), 39-62. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 67 Racicot, R. (2016). The effect of multimedia writing support software on written productivity. Journal of Occupational Therapy, Schools, & Early Intervention, 9(1), 99-123. Sessions, L., Kang, M. O., & Womack, S. (2016). The neglected "R": Improving writing instruction through iPad apps. Techtrends: Linking Research and Practice To Improve Learning, 60(3), 218-225. Spooner, F., Kemp-Inman, A., Ahlgrim-Delzell, L., Wood, L., & Ley Davis, L. (2015). Generalization of literacy skills through portable technology for students with severe disabilities. Research and Practice for Persons With Severe Disabilities, 40(1), 52-70. Suhr, K. A., Hernandez, D. A., Grimes, D., & Warschauer, M. (2010). Laptops and fourth-grade literacy: Assisting the jump over the fourth-grade slump. Journal of Technology, Learning, and Assessment, 9(5). Sung, Y., Chang, K., & Liu, T. (2016). The effects of integrating mobile devices with teaching and learning on students' learning performance: A meta-analysis and research synthesis. Computers & Education, 94252-275. doi:10.1016/j.compedu.2015.11.008 Trochim, W.M.K. (2006). Hybrid experimental designs. In Research methods knowledge base. Retrieved from www.socialresearchmethods.net/kb/exphybrd.php Wollak, B. A., & Koppenhaver, D. A. (2011). Developing technology-supported, evidence-based writing instruction for adolescents with significant writing disabilities. Assistive Technology Outcomes and Benefits, 7(1), 1-23. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 68 Wollscheid, S., Sjaastad, J., Tømte, C., & Løver, N. (2016). The effect of pen and paper or tablet computer on early writing–a pilot study. Computers & Education, 98, 70-80. Wu, W. H., Wu, Y. C. J., Chen, C. Y., Kao, H. Y., Lin, C. H., & Huang, S. H. (2012). Review of trends from mobile learning studies: A meta-analysis. Computers & Education, 59(2), 817-827. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 69 APPENDIX A: Participant Raw Data Group Grade Gender Diagnosed Disabilities FAR SS Far PR Writing Screener Score Screener Combined Score F33 A 3 M LD, juvenile arthritis 59 0.3 0 59 B76 A 2 M ADHD, Oppositional Defiant Disorder 69 2 0 69 B73 A 2 M Developmental Coordination Disorder 72 3 0 72 E31 A 5 M LD 81 10 5 124.7 F19 A 3 M None 83 13 6 135.44 E88 A 5 M None 93 32 7 154.18 D37 A 6 F Microcephaly, MID 83 13 9 161.66 C48 A 7 F None 78 7 10 165.4 E14 A 5 F None 98 45 9 176.66 E81 A 5 F LD 99 47 9 177.66 A26 A 4 F FASD, MID, ADHD, Language Disorder 78 7 7 139.18 D29 B 6 M Non-verbal LD, Sensory Processing Disorder, Anxiety Disorder NOS, Mood Disorder NOS, Expressive Language Disorder, ADHD, Stereotypic Movement Disorder NOS 70 2 9 148.66 F52 B 3 M LD, Developmental Coordination Disorder 67 1 0 67 B12 B 2 M ADHD, FASD, Sensory Processing Disorder, 69 2 0 69 Participant Attachment Disorder, GDD E64 B 5 M FASD, GDD, MID, ADHD 62 1 5 105.7 B23 B 2 F None 64 1 7 125.18 E74 B 5 F None 83 13 8 152.92 E96 B 5 F None 83 13 9 161.66 A65 B 4 F None 94 34 8 163.92 B42 B 2 F None 84 14 10 171.4 C56 B 7 M Oppositional Defiant Disorder, ADHD 90 25 10 177.4 E92 B 5 F LD 98 45 10 185.4 Legend: LD- Learning Disability, ADHD- Attention Deficit Hyperactivity Disorder, FASD- Fetal Alcohol Spectrum Disorder, MID-Mild Intellectual Disability, GDD-Global Developmental Delay ASSISTIVE TECHNOLOGY TO ENHANCE WRITING APPENDIX B: Writing Quality Assessment Instructions Assessment – Teacher Instructions for Grades 2-7 1. Give each child a copy of the same motivating picture prompt so that all students are writing for the same purpose. 2. Set clear timelines an offer friendly reminders throughout the writing session:  Planning time: 10 minutes (in small groups of 3-5) Talk about the following: a) Look carefully at the picture and talk about what you think may be happening b) What may have happened before this point? c) What could happen after? d) Think of a story you could make with at least two people, things or animals that speak.  Brainstorm ideas for your writing  Write an outline of a beginning, middle and end of your story  Writing time: 30 minutes  Revision time: 5-10 minutes 70 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING APPENDIX C: Writing Quality Assessment Grade Level Rubrics 71 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 72 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 73 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 74 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 75 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 76 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING APPENDIX D: Writing Attitude Survey Sample Page 77 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING APPENDIX E: Screenshots of the Clicker Docs Application by Crick Software Word Processor Screen Word Prediction 78 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING Word Banks Superkeys: Accessible Keyboard 79 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 80 APPENDIX F: Assistive Technology Intervention Lesson Plans Assistive Technology Intervention Lesson Plans (25 minutes each) *Teacher models lessons on own iPad using Apple TV and projector Lesson 1: Accessibility Features/ Notes Application A. Siri: -Ask Students if they know what Siri is. Explain to students that Siri is a personal digital assistant. This means that Siri is software or a computer program that is designed to help the user and answer questions. -You can use Siri for many different things. -Model for students how to access Siri by holding down the home button. Ask “Siri, What can you do?” And discuss some of the answers with the students. -Show students how to check if their volume is up and that the iPad is not on mute. -Some things that may be helpful when writing are 1.) Spelling: Model “How do you spell beautiful?” Have a few children ask Siri a spelling question of their choice on their iPads. 2.) Dictionary: Model “What does diminish mean?” Have a few children try a definition of their choice on their iPads. 3.) Thesaurus: “What is another word for big” Have children ask Siri a question of their choice on their iPads. 4.) Question and answer: Ask Siri information questions. This could be helpful in nonfiction writing. 5.) Commands: You can have Siri open applications. Have students “Open notes” B. Accessibility Features using the Notes Application: -We will be using the notes application to practice an accessibility feature called ‘Dictation’. Dictation is a way of doing something called speech to text. On most tablet computers this is a built-in feature that will type what you say. This can be really helpful if you find spelling or writing down your ideas fast enough difficult. It is usually build into the tablet computer keyboard. But the application we will mostly be learning to use, Clicker Docs has its own keyboard, so if you want to use Dictation we will learn how to copy and paste from the Notes application. First, we will practice Dictation. (Model for students) 1.) In Notes, tap on the bottom of the screen. This should make your keyboard appear. 2.) Find the button that looks like a microphone next to the space bar and press it. 3.) Speak your sentence. Hit the done button when you are done recording. Dictation will only record 30 seconds at a time, so it is better to use quick sentences. You need to say the word to use punctuation. Model “I love teaching students to do new things” and “What is your favorite color?” Have students try these sentences and then answer the last one using Dictation. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 81 Lesson 2: Split Screen/ Intro to Clicker Docs Review Dictation on notes. - Spend approximately 5 minutes writing a few sentences. It could be a message to a friend or something different. A. Split Screen and Copy/Paste from Notes Now we will learn how to copy our message into the program we will be using… Clicker Docs 1.) 2.) 3.) 4.) 5.) 6.) Open Clicker Docs. Use Siri or press the home button, find the Clicker Docs button and press it. In Clicker Docs, press the file folder or + button to make a new document. Name your document by typing or using Dictation Click done Click on your document name Now we will learn how to do split screen. This will allow you to have 2 programs open at once. Touch the right hand side of your screen and swipe quickly right to left from the edge. Notes may come up, or you may need to swipe down to see the notes app. 7.) Your sentences should appear. Hold your finger down on top of the text. Select the text by dragging the ends to highlight everything you want to select and then press the copy button or press select all if it comes up. 8.) Now touch the Clicker Docs application. Hold your finger down until a menu appears. Touch the paste button. 9.) Practice using Dictation to write a message in notes and copying and pasting into Clicker Docs for a few minutes. B. Introduction to Clicker Docs Explore keyboard buttons- Pointing to the keyboard on the projector screen, have children take note of: 1.) 2.) 3.) 4.) 5.) 6.) 7.) 8.) 9.) the letters for typing the backspace button the up arrow for making letters uppercase the return button the 123 button for getting to the numbers keyboard single arrows for moving left and right in a document Double arrows for moving left and right by whole word in the document ABC button for getting back to the letters keyboard Hide the keyboard button Introduction to Settings: - Demonstrate where the gear is at the top of the document. Have children find and press it on their iPad. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 82 -Explain that ‘Document’ is where you can change your font, font size, background color and text color. Have children find these settings, but explain that these are style choices that we are not changing right now. Click on the ‘Settings’ button to go back. -Explain that one of the features of Clicker Docs is that it will read your sentences or words back to you. We can activate this by clicking on ‘Speech’. The first part is voice. We can choose a computer voice that is close to our own. If you click on ‘Voice’ a number of names and flags come up. The voice that is closest to ours are the ones from the United States, so select Ella if you are a girl and Josh if you are a boy. Click on ‘Speech’ to return to the menu. ‘Speed’ is how fast the computer reads your words or sentences. For now it should be set at medium but you can change this if you find it is too fast or slow. ‘Highlight color’ is the color the computer will highlight your words in. This is set at red which is fine for now. The bottom part gives us the option of having the computer speak each letter, each word, or speak each sentence. For now turn on ‘Speak each sentence’ by sliding the toggle button. It will turn green to let you know it is on. Click ‘Settings’ to go back to the menu. Now try out the speech setting by holding your finger for a few sections somewhere on the white screen part of your menu. A black menu should pop up. Click on the ‘speak all’ button to have your writing read to you. (*Researcher and EA’s assist students to implement features learned). Lesson 3: Clicker Docs settings/Word Banks Review: review keyboard keys. Ensure settings for voice are set on each iPad. Go over ‘Speak each word’ and ‘Speak each letter’ options by modelling on the teacher iPad. Children may choose which settings they feel will work for them. Discourage from using all three. Have children type a sentence to try their settings. Review how to access ‘Speak all’ setting. A. Clicker Docs Settings-spell checker and predictor Spell Checker: Explain that Clicker Docs also has a spell checker feature. Have students click on ‘settings’, then ‘spell checker and predictor’ and then turn ‘Spell checker’ on by sliding the toggle. It should turn green. When this is turned on, any words that are spelled incorrectly will be underlined with a red dashed line. Spell the word ‘have’ as ‘hav’ on the teacher iPad. Note how the red line appears. Have students do this on their iPad. If you click on that word options will appear to correct it (model on iPad). Maybe you don’t know what the words it gives you says. This is when it is helpful to use the speaker button at the top of the keyboard (point on the screen). If you press the speaker button first and then click on the word, it will say it for you. You need to turn the speaker on for each word. Model this on the teacher iPad. Once ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 83 you find the word you are looking for, click on it and it will replace the word in your writing. The red line should disappear. Have students try on their iPad. Predictor: Explain that Clicker Docs also has something called a word predictor. This means that the computer can suggest words that you might want to use once you start typing. Some of you might have seen something similar if you have ever typed on a cellphone, or have seen this on your parent’s cellphone. To turn on ‘Predictor” we need to go back to settings. Click on ‘Spell Checker and Predictor’. Next click on ‘predictor’. Slide the toggle button to turn it on. It should turn green. The next area is sounds like prediction. This means that the computer will predict words that sound similar to what you are typing. If you click on this you can select ‘off’, ‘low’, ‘medium’, or ‘high.’ For this, we are just going to select ‘medium’ for now. Click on ‘predictor’ to go back to the menu. The next option is ‘predict next word.’ If you turn this on, it will predict words that might come logically next in your sentence. Slide the toggle to turn it on. The next setting is ‘database size’. This is how many words that your computer is selecting from. I have mine set at large. If you click on ‘Back’ there is one more setting we can adjust. ‘Display’ is how you change what the word choices look like on your screen. First of all you can choose how many words are displayed. This can range from 3-8 at a time. If you get overwhelmed by reading you may want to choose less words. If you want more options you should choose more words. I’m going to set mine at 8. You can change the number by pressing the – or + button. The last setting is how big the words are that appear on your screen. If you find it hard sometimes to click on things that are small you are going to want to set your words larger. This will also take up more room on your screen. Show students what ‘Very large’, ‘Large’, ‘Medium’, and ‘small’ looks like on the projector. Have them choose a size that they prefer. Practice writing a few sentences using these features. (*Researcher and EA’s assist students to implement features learned). Lesson 4: Clicker Docs Settings: Superkeys and Word Banks A. Superkeys: Clicker Docs also has the option of changing the keyboard so that the buttons become larger. This can help if you find it difficult to tap the right keys when you are typing. To turn on Superkeys we need to go to ‘Settings’. Next we go down to ‘Accessibility’. At the top you tap on ‘Superkeys’ and slide the toggle button to turn it on. Then you can decide whether you want the keyboard to close automatically and after how many seconds, or whether you want to control that. Model the close automatically feature. Model and without. Have students turn on Superkeys and practice writing a sentence or two. They can decide to leave it on if it is helpful or to turn it back off. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 84 B. Word Banks: Clicker Docs also has word banks that give ideas for vocabulary on certain themes. Word banks can be created or sent to you by a teacher. For this research study we will not be making our own word banks but we will be using some that I send you. First we need to learn how to open a word bank. 1.) Click on the file folder in the top right hand corner. If you have any word banks available to you. They will show up in the list of documents. They have a rectangle icon. 2.) You can also click on ‘Learning Grid downloads’. This will give you access to any downloaded word banks. I have downloaded one called “My Spare Time’ which I will show you right now and then share with you later. Click on the word bank to open it. 3.) Your word bank should then come up instead of your keyboard. You can switch back and forth between word banks and keyboard by clicking on either of the icons (model for students). 4.) Model how to use a word bank for the students. There are categories at the bottom. This learning grid is my spare time, so it has words for spare time activities at home, outings words, reading words, arts and crafts words, music words, sports words, and outdoors words. You can navigate back and forth by pressing the arrow keys. Click on one of the titles to select words for that category. Model this for students. 5.) To use some of the words in your writing, simply click on the word. You can use the speaker to have the word read to you and the icons to switch back and forth between the keyboard and word bank. Now we will attempt to download this word bank onto your iPad and start our first writing assignment. We will get a chance to practice using word banks and the other features that we have learned so far. 1.) First, let’s make a new document so that we can write about our free time in. Press on the file folder and use the + sign. Click on Document. Name your document something related to free time. 2.) Now I am going to air drop the word bank onto your iPad. The first thing you need to do is turn airdrop on. To do this swipe your finger up from the very bottom of the screen. You should see a button called “airdrop”. Click on this button. Then click on ‘everyone’. Now your iPad should be ready to receive the word bank. 3.) After I send it to your iPad you should be able to open it. Go ahead and use the word bank to write about what you do in your free time. (*Researcher and EA’s assist students to implement features learned). Lesson 5: Pictures Last time we learned about Superkeys and word banks and we started writing about our free time. Today we are going to learn how to insert a photo. A. Pictures It is possible to put pictures in a clicker Doc and there are several ways you can do this. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 85 1. Camera roll: -You can take a picture with your iPad to put in your writing (Have students do this with you as you demo). -Swipe your finger up from the bottom and press the camera button. - Aim your iPad and take a photo by clicking on the circle button. -Now open your photos by swiping from right to left and then down from the top. -Scroll down until you see photos and open it up. -Click on the photo you would like to add. At the bottom of that photo click on the box with the arrow on the top and select copy. -Hold your finger down on clicker docs in the spot you want your picture and select paste. 2. Internet photos: -You can copy and paste a picture from the internet in the same way. Open up Safari in your split screen. Search for a photo and copy and paste. 3. Drawing: -Draw something in notes. Open your notes keyboard and select the squiggle at the top. Draw something using the tools, select the box with the arrow at the top and copy and paste. (*Researcher and EA’s assist students to implement features learned). Lesson 6: Writing/Sharing We are going to finish our free time writing assignment and spend a few minutes sharing with each other. B. Practice/ Writing Spend no more than a few minutes selecting a photo to insert in your free time assignment and then finish off your assignment. We'll spend the last 7 minutes sharing our writing with each other. (*Researcher and EA’s assist students to implement features learned). C. Sharing our writing We are going to learn how to share our writing by mirroring to the projector: - Swipe up from the bottom and select airplay: mirroring. -Then when I call on you to share you can select Apple TV. Type in the 4 digit code it gives you and you that will mirror your iPad and block others out. -You can share your story by reading it or selecting speak all on Clicker Docs. Make sure your volume is all the way up. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 86 Lesson 7: Planning for writing (Beginning/Middle/End) Today we are going to spend some time learning how to plan for writing using an iPad. We can plan using clicker docs, or using another app called Popplet. A. Planning on clicker docs: Beginning, middle, End (Have students follow along on their iPad as you model) - Open a new document and name it Beginning, Middle, End template (or something similar).Type the words Beginning, Middle, End somewhere on your screen or even B-M-E. Add a few ideas under each heading. Model an example for students. B. Planning on Popplet: Beginning, Middle, End (Have students follow along on their iPad as you model. Click on the ? to demonstrate to students) Popplet is an application that allows us to plan for writing by making a graphic organizer, which is like a web. -Click on the words at the top left hand corner of the screen to name your popplet page. -Click on the Gear and then make new popple -Use the arrows on either of the corners to make your popple bigger or smaller. -Tap on the T to type (you can use Dictation). Label it Beginning -Click on the circle on the right side. This makes a connected idea. Label it middle. -Click on the circle on the right side again to make another idea. Label it end. -Make new popples down from those to make boxes to type ideas, or type directly in the boxes. -Use your fingers on the screen to move around your popplet. C. Practice with Planning Use clicker docs or Popplet to plan for a story that takes place in our school. If you would like to use a picture for inspiration you can spend a few minutes taking, drawing or selecting a picture first. Open my school word bank that I will send to your iPad and begin writing a story using a school as a setting. (*Researcher and EA’s assist students to implement features learned). Lesson 8: Planning/Writing Finish planning for your story that takes place in a school. Go ahead and use Clicker Docs to write your story. (*Researcher and EA’s assist students to implement features learned). ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 87 Lesson 9: Printing/ Editing Today we are going to spend some time talking about editing our work. We will edit our writing from last time and learn how to print our documents. Printing First we will make sure that everyone knows how to print from our iPad. Start by opening Clicker Docs and open your free time writing assignment. We are going to try and print this document. It’s okay if it is not finished. The first thing you need to decide is what way you would like your page to print. If you are holding your iPad horizontally, your page will print in landscape mode. If you are holding your iPad vertically, your page will print in portrait mode. First click on the paper icon with the folded corner on the top left hand corner of the screen. A ‘Document’ menu should come up. Click on the ‘print’ button on this menu. Next click on ‘select Printer’. The iPad will search for Air printers in our school. Our air printer is the secretary’s printer so we will select this one. Go ahead and click print and we’ll send someone to the printer to make sure that it worked. Editing There are a number of things we can do to make sure our writing makes sense and sounds good. (Write these ideas on the whiteboard) 1.) Spell check- check for any words underlined in red and see if you can correct them. 2.) Text-to-speech- While you are writing, make sure your headphones are plugged in. Every time you type a period, your sentence will be read back to you. Listen carefully to make sure it sounds right. Fix any words that don’t sound right, add words that you might have missed, and delete any extra words. 3.) When you are finished writing, hold down your finger on the screen to select the “speak all” button. Make sure your story makes sense and sounds right. 4.) Interesting words- If any of your words sound boring, try thinking of a more exciting word, or use a word bank to help you. Finish writing your school story. We will only have a few minutes to do this. Then go ahead and edit the story to make sure it makes sense and sounds good. You can read it to someone beside you if you have time. We will print these stories at the end of the session. (*Researcher and EA’s assist students to implement features learned). Lesson 10: Word banks to support writing/Writing Process Today I am going to show you a few more word banks to support your writing, and show you how you can use more than one word bank at a time. Then we will practice writing a quick story. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 88 First open up any of your documents. To get to a word bank you click on the file folder at the top. You can go back and forth between word banks in the same document. I’ll send you some word banks as we go along to add to your word bank library. Make sure your airdrop is turned to everyone so that you can receive them. First I’ll send you ‘Traditional Story words’. This one has words for things like characters, setting, and some nouns, adjectives, and transition words. If you were writing a make-believe story, this one has great words to help. Now let’s say you were writing a story and you wanted to use a word that you know is in the free time word bank. You can do this by clicking on the folder at the top and then clicking on ‘my free time’ word bank. You can go back and forth as often as you like. I am going to send you a few more word banks.          ‘Story setting adjectives’ is helpful for thinking of really interesting words to describe the settings in your stories. ‘Spooky words A-Z’ is helpful if you want to write a story that is a bit scarier. These words are arranged alphabetically like a dictionary so it is helpful to know the first sound of the word. ‘House words’ are helpful for any words about a house for a setting. ‘Good character, bad character’ gives words that describe different characters. ‘Describe the weather A-Z’ gives some great words for talking about the weather in your setting. ‘Character words’ gives great descriptive words for describing a character- what he or she looks like. ‘Alternative adjectives’-gives more interesting words for words that are more boring. This is kind of like the triple scoop words we use in class. ‘Adjectives A-Z’ is a dictionary of great describing words to help make your writing more interesting. ‘Alternatives to said’ gives more interesting words that you can use instead of the word said. I would like you to spend a few minutes writing sentences using these word banks so that you become more familiar with them. The sentences do not have to make a story. Just try and make a few interesting sentences. (*Researcher and EA’s assist students to implement features learned). Lesson 11: Writing Process For our last 2 lessons we will be practicing the whole writing process in order to put together everything that we have learned. Today I will be showing you a picture prompt. We will plan for writing using one of our beginning, middle, end organizers and discussing with a small group. We will begin writing a story. Next class we will finish the story, edit, and share with each other. Here is a picture prompt for writing today. We’ll take 10 minutes to plan. e) In a group of 3-5 discuss Look carefully at the picture and talk about what you think may be happening ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 89 f) What may have happened before this point? g) What could happen after? h) Think of a story you could make with at least two people, things or animals that speak. i) Write down some of your ideas in your BME document or popplet. Remember you only have a few minutes so just write down key words, no, sentences. Create a new document and give it a title. Go ahead and begin writing your story. We will take the next 10 minutes for writing and then we will have 15 minutes more next class. Remember that you can use any of the word banks, you can use Dictation speech to text with notes in split screen, word predictor, Siri for spelling, or anything else that we have learned in our sessions. (*Researcher and EA’s assist students to implement features learned). Lesson 12: Writing Process continued Today is our last session! We are going to take 15 minutes to finish writing our stories. Then use some time for editing. Remember to check for spelling and interesting words, and make sure your sentences make sense using ‘speak all’. After editing share your story with someone at your table, and print them. (*Researcher and EA’s assist students to implement features learned). Thanks to everyone for participating in the iPad sessions. I hope you enjoyed them. I will see you 2 more times for an assessment. After this next assessment your iPad will go to the next group, so unfortunately you will not be able to use them in class anymore while the other group is learning. I hope you have enjoyed your time with the iPad and if you have any questions, you can feel free to come and speak to me. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING APPENDIX G: Planning for Writing Assessment Page and Writing Paper Writing Name: ________________________________ Date: ___________________ Grade: _______ In a small group talk about the picture and decide what you think…  What is happening in the picture?  What may have happened before this?  What could happen after?  Describe the characters involved? Brainstorm ideas for your story: 90 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 91 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 92 APPENDIX H : SPSS Data Analysis for Question 1 GLM WQComp_PT1 WQComp_PT2 BY Group WITH WQComp_Pre /WSFACTOR=Time 2 Polynomial /METHOD=SSTYPE(3) /PLOT=PROFILE(Time*Group) /EMMEANS=TABLES(Group*Time) WITH(WQComp_Pre=MEAN) /PRINT=DESCRIPTIVE ETASQ HOMOGENEITY /CRITERIA=ALPHA(.05) /WSDESIGN=Time /DESIGN=WQComp_Pre Group. [DataSet1] C:\Users\hstacesmith\Google Drive\TWU Masters\capstone project\data\Masters thesis results.sav General Linear Model Within-Subjects Factors Measure: MEASURE_1 Dependent Time Variable 1 WQComp_PT1 2 WQComp_PT2 Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Group Mean Std. Deviation N Writing Quality Composite Group A 8.3636 1.89856 11 Post-test 1 Group B 7.0909 1.80025 11 Total 7.7273 1.91937 22 Writing Quality Compostie Group A 7.2727 1.29158 11 Post-test 2 Group B 8.7273 2.30612 11 Total 8.0000 1.97001 22 Box's Test of Equality of Covariance Matricesa ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 93 Box's M 6.081 F 1.807 df1 3 df2 72000.000 Sig. .143 Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups. a. Design: Intercept + WQComp_Pre + Group Within Subjects Design: Time Tests of Within-Subjects Effects Measure: MEASURE_1 Type III Sum Source Time of Squares Partial Eta df Mean Square F Sig. Squared Sphericity Assumed 5.621 1 5.621 4.135 .056 .179 Greenhouse-Geisser 5.621 1.000 5.621 4.135 .056 .179 Huynh-Feldt 5.621 1.000 5.621 4.135 .056 .179 Lower-bound 5.621 1.000 5.621 4.135 .056 .179 Time * Sphericity Assumed 6.900 1 6.900 5.076 .036 .211 WQComp_Pre Greenhouse-Geisser 6.900 1.000 6.900 5.076 .036 .211 Huynh-Feldt 6.900 1.000 6.900 5.076 .036 .211 Lower-bound 6.900 1.000 6.900 5.076 .036 .211 Sphericity Assumed 21.748 1 21.748 15.999 .001 .457 Greenhouse-Geisser 21.748 1.000 21.748 15.999 .001 .457 Huynh-Feldt 21.748 1.000 21.748 15.999 .001 .457 Lower-bound 21.748 1.000 21.748 15.999 .001 .457 Sphericity Assumed 25.828 19 1.359 Greenhouse-Geisser 25.828 19.000 1.359 Huynh-Feldt 25.828 19.000 1.359 Lower-bound 25.828 19.000 1.359 Time * Group Error(Time) Tests of Within-Subjects Contrasts Measure: MEASURE_1 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 94 Type III Sum of Squares Partial Eta Source Time df Mean Square F Sig. Time Linear 5.621 1 5.621 4.135 .056 .179 Time * WQComp_Pre Linear 6.900 1 6.900 5.076 .036 .211 Time * Group Linear 21.748 1 21.748 15.999 .001 .457 Error(Time) Linear 25.828 19 1.359 Levene's Test of Equality of Error Variancesa F Writing Quality Composite Post-test 1 Writing Quality Compostie Post-test 2 df1 df2 Sig. .293 1 20 .594 .527 1 20 .476 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WQComp_Pre + Group Within Subjects Design: Time Tests of Between-Subjects Effects Measure: MEASURE_1 Transformed Variable: Average Type III Sum of Source Squares Partial Eta df Mean Square F Sig. Squared Intercept 16.431 1 16.431 6.022 .024 .241 WQComp_Pre 53.750 1 53.750 19.700 .000 .509 .510 1 .510 .187 .670 .010 51.841 19 2.728 Group Error Squared ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 95 Estimated Marginal Means Group * Time Measure: MEASURE_1 95% Confidence Interval Group Time Group A 1 8.324a .524 7.226 9.421 2 7.188a .312 6.535 7.841 1 7.131a .524 6.033 8.229 2 8.812a .312 8.159 9.465 Group B a. Mean Std. Error Lower Bound Upper Bound Covariates appearing in the model are evaluated at the following values: Writing Quality Composite Pre-test = 7.3182. UNIANOVA WQComp_Pre BY Group /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /CRITERIA=ALPHA(0.05) /DESIGN=Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Tests of Between-Subjects Effects Dependent Variable: Writing Quality Composite Pre-test Type III Sum of Source Squares df Mean Square F Sig. .182a 1 .182 .064 .803 1178.227 1 1178.227 412.755 .000 .182 1 .182 .064 .803 Error 57.091 20 2.855 Total 1235.500 22 57.273 21 Corrected Model Intercept Group Corrected Total a. R Squared = .003 (Adjusted R Squared = -.047) ASSISTIVE TECHNOLOGY TO ENHANCE WRITING UNIANOVA WQComp_PT1 BY Group WITH WQComp_Pre /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /EMMEANS=TABLES(OVERALL) WITH(WQComp_Pre=MEAN) /EMMEANS=TABLES(Group) WITH(WQComp_Pre=MEAN) COMPARE ADJ(LSD) /PRINT=ETASQ HOMOGENEITY DESCRIPTIVE /CRITERIA=ALPHA(.05) /DESIGN=WQComp_Pre Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Dependent Variable: Writing Quality Composite Posttest 1 Group Mean Std. Deviation N Group A 8.3636 1.89856 11 Group B 7.0909 1.80025 11 Total 7.7273 1.91937 22 Levene's Test of Equality of Error Variancesa Dependent Variable: Writing Quality Composite Post-test 1 F df1 .293 df2 1 Sig. 20 .594 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WQComp_Pre + Group 96 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 97 Tests of Between-Subjects Effects Dependent Variable: Writing Quality Composite Post-test 1 Type III Sum of Source Partial Eta Squares df Mean Square F Sig. Squared 19.976a 2 9.988 3.307 .059 .258 Intercept 20.636 1 20.636 6.832 .017 .264 WQComp_Pre 11.067 1 11.067 3.664 .071 .162 Group 7.799 1 7.799 2.582 .125 .120 Error 57.387 19 3.020 Total 1391.000 22 77.364 21 Corrected Model Corrected Total a. R Squared = .258 (Adjusted R Squared = .180) Estimated Marginal Means 1. Grand Mean Dependent Variable: Writing Quality Composite Post-test 1 95% Confidence Interval Mean 7.727a Std. Error Lower Bound .371 Upper Bound 6.952 8.503 a. Covariates appearing in the model are evaluated at the following values: Writing Quality Composite Pre-test = 7.3182. 2. Group Estimates Dependent Variable: Writing Quality Composite Post-test 1 95% Confidence Interval Group Mean Std. Error Lower Bound Upper Bound Group A 8.324a .524 7.226 9.421 Group B 7.131a .524 6.033 8.229 a. Covariates appearing in the model are evaluated at the following values: Writing Quality Composite Pre-test = 7.3182. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 98 Pairwise Comparisons Dependent Variable: Writing Quality Composite Post-test 1 95% Confidence Interval for Differencea Mean Difference (I-J) Std. Error Sig.a (I) Group (J) Group Lower Bound Upper Bound Group A Group B 1.193 .742 .125 -.361 2.746 Group B Group A -1.193 .742 .125 -2.746 .361 Based on estimated marginal means a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Dependent Variable: Writing Quality Composite Post-test 1 Partial Eta Sum of Squares Contrast Error df Mean Square 7.799 1 7.799 57.387 19 3.020 F 2.582 Sig. Squared .125 The F tests the effect of Group. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. UNIANOVA WQComp_PT2 BY Group WITH WQComp_Pre /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /EMMEANS=TABLES(OVERALL) WITH(WQComp_Pre=MEAN) /EMMEANS=TABLES(Group) WITH(WQComp_Pre=MEAN) COMPARE ADJ(LSD) /PRINT=ETASQ HOMOGENEITY DESCRIPTIVE /CRITERIA=ALPHA(.05) /DESIGN=WQComp_Pre Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Dependent Variable: Writing Quality Compostie Posttest 2 .120 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING Group Mean Std. Deviation 99 N Group A 7.2727 1.29158 11 Group B 8.7273 2.30612 11 Total 8.0000 1.97001 22 Levene's Test of Equality of Error Variancesa Dependent Variable: Writing Quality Compostie Post-test 2 F df1 .527 df2 1 Sig. 20 .476 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WQComp_Pre + Group Tests of Between-Subjects Effects Dependent Variable: Writing Quality Compostie Post-test 2 Type III Sum of Source Partial Eta Squares df Mean Square F Sig. Squared 61.219a 2 30.610 28.676 .000 .751 1.416 1 1.416 1.326 .264 .065 WQComp_Pre 49.583 1 49.583 46.451 .000 .710 Group 14.459 1 14.459 13.546 .002 .416 Error 20.281 19 1.067 Total 1489.500 22 81.500 21 Corrected Model Intercept Corrected Total a. R Squared = .751 (Adjusted R Squared = .725) Estimated Marginal Means 1. Grand Mean Dependent Variable: Writing Quality Compostie Post-test 2 95% Confidence Interval Mean 8.000a Std. Error .220 Lower Bound Upper Bound 7.539 8.461 a. Covariates appearing in the model are evaluated at the following values: Writing Quality Composite Pre-test = 7.3182. ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 100 2. Group Estimates Dependent Variable: Writing Quality Compostie Post-test 2 95% Confidence Interval Group Mean Std. Error Lower Bound Upper Bound Group A 7.188a .312 6.535 7.841 Group B 8.812a .312 8.159 9.465 a. Covariates appearing in the model are evaluated at the following values: Writing Quality Composite Pre-test = 7.3182. Pairwise Comparisons Dependent Variable: Writing Quality Compostie Post-test 2 95% Confidence Interval for Differenceb Mean Difference (I-J) Std. Error Sig.b (I) Group (J) Group Lower Bound Upper Bound Group A Group B -1.624* .441 .002 -2.548 -.700 Group B Group A 1.624* .441 .002 .700 2.548 Based on estimated marginal means *. The mean difference is significant at the .05 level. b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Dependent Variable: Writing Quality Compostie Post-test 2 Partial Eta Sum of Squares df Mean Square Contrast 14.459 1 14.459 Error 20.281 19 1.067 F 13.546 Sig. Squared .002 The F tests the effect of Group. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. .416 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 101 Appendix I : SPSS Data Analysis for Question 2 GLM WC_PT1 WC_PT2 BY Group WITH WC_Pre /WSFACTOR=time 2 Polynomial /METHOD=SSTYPE(3) /PLOT=PROFILE(time*Group) /EMMEANS=TABLES(Group) WITH(WC_Pre=MEAN)COMPARE ADJ(LSD) /EMMEANS=TABLES(time) WITH(WC_Pre=MEAN)COMPARE ADJ(LSD) /EMMEANS=TABLES(Group*time) WITH(WC_Pre=MEAN) /PRINT=DESCRIPTIVE ETASQ HOMOGENEITY /CRITERIA=ALPHA(.05) /WSDESIGN=time /DESIGN=WC_Pre Group. General Linear Model Within-Subjects Factors Measure: MEASURE_1 Dependent time Variable 1 WC_PT1 2 WC_PT2 Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Group Word Count Post-test 1 Word Count Post-test 2 Mean Std. Deviation N Group A 98.00 70.529 11 Group B 165.45 134.424 11 Total 131.73 110.296 22 Group A 147.55 65.343 11 Group B 114.18 83.242 11 Total 130.86 74.995 22 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 102 Box's Test of Equality of Covariance Matricesa Box's M 10.234 F 3.042 df1 3 df2 72000.000 Sig. .028 Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups. a. Design: Intercept + WC_Pre + Group Within Subjects Design: time Multivariate Testsa Partial Eta Effect time time * WC_Pre time * Group Value F Hypothesis df Error df Sig. Squared b 1.000 19.000 .576 .017 Pillai's Trace .017 .324 Wilks' Lambda .983 .324b 1.000 19.000 .576 .017 Hotelling's Trace .017 .324b 1.000 19.000 .576 .017 Roy's Largest Root .017 .324b 1.000 19.000 .576 .017 .025 .488b 1.000 19.000 .493 .025 Wilks' Lambda .975 .488b 1.000 19.000 .493 .025 Hotelling's Trace .026 .488b 1.000 19.000 .493 .025 Roy's Largest Root .026 .488b 1.000 19.000 .493 .025 Pillai's Trace .189 4.441b 1.000 19.000 .049 .189 Wilks' Lambda .811 4.441b 1.000 19.000 .049 .189 Hotelling's Trace .234 4.441b 1.000 19.000 .049 .189 Roy's Largest Root .234 4.441b 1.000 19.000 .049 .189 Pillai's Trace a. Design: Intercept + WC_Pre + Group Within Subjects Design: time b. Exact statistic Tests of Within-Subjects Effects ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 103 Measure: MEASURE_1 Type III Sum of Source Partial Eta Squares time time * WC_Pre time * Group Error(time) df Mean Square F Sig. Squared Sphericity Assumed 2054.026 1 2054.026 .324 .576 .017 Greenhouse-Geisser 2054.026 1.000 2054.026 .324 .576 .017 Huynh-Feldt 2054.026 1.000 2054.026 .324 .576 .017 Lower-bound 2054.026 1.000 2054.026 .324 .576 .017 Sphericity Assumed 3088.664 1 3088.664 .488 .493 .025 Greenhouse-Geisser 3088.664 1.000 3088.664 .488 .493 .025 Huynh-Feldt 3088.664 1.000 3088.664 .488 .493 .025 Lower-bound 3088.664 1.000 3088.664 .488 .493 .025 Sphericity Assumed 28125.393 1 28125.393 4.441 .049 .189 Greenhouse-Geisser 28125.393 1.000 28125.393 4.441 .049 .189 Huynh-Feldt 28125.393 1.000 28125.393 4.441 .049 .189 Lower-bound 28125.393 1.000 28125.393 4.441 .049 .189 Sphericity Assumed 120316.791 19 6332.463 Greenhouse-Geisser 120316.791 19.000 6332.463 Huynh-Feldt 120316.791 19.000 6332.463 Lower-bound 120316.791 19.000 6332.463 Tests of Within-Subjects Contrasts Measure: MEASURE_1 Type III Sum of Partial Eta Source time Squares df Mean Square time Linear 2054.026 1 2054.026 .324 .576 .017 time * WC_Pre Linear 3088.664 1 3088.664 .488 .493 .025 time * Group Linear 28125.393 1 28125.393 4.441 .049 .189 Error(time) Linear 120316.791 19 6332.463 Levene's Test of Equality of Error Variancesa F df1 df2 Sig. Word Count Post-test 1 4.344 1 20 .050 Word Count Post-test 2 2.032 1 20 .169 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WC_Pre + Group Within Subjects Design: time F Sig. Squared ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 104 Tests of Between-Subjects Effects Measure: MEASURE_1 Transformed Variable: Average Type III Sum of Source Partial Eta Squares df Mean Square F Sig. Squared Intercept 22911.179 1 22911.179 5.593 .029 .227 WC_Pre 141200.479 1 141200.479 34.472 .000 .645 Group 3610.249 1 3610.249 .881 .360 .044 Error 77825.158 19 4096.061 3. Group * time Measure: MEASURE_1 95% Confidence Interval Group time Group A 1 97.385a 25.642 43.715 151.054 2 147.089a 17.046 111.411 182.767 1 166.070a 25.642 112.400 219.739 2 114.638a 17.046 78.960 150.317 Group B Mean Std. Error Lower Bound Upper Bound a. Covariates appearing in the model are evaluated at the following values: Word Count Pre-test = 150.73. UNIANOVA WC_Pre BY Group /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /CRITERIA=ALPHA(.05) /DESIGN=Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Tests of Between-Subjects Effects Dependent Variable: Word Count Pre-test ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 105 Type III Sum of Source Squares df Mean Square F Sig. 18.182a 1 18.182 .002 .967 499811.636 1 499811.636 49.199 .000 18.182 1 18.182 .002 .967 Error 203178.182 20 10158.909 Total 703008.000 22 Corrected Total 203196.364 21 Corrected Model Intercept Group a. R Squared = .000 (Adjusted R Squared = -.050) UNIANOVA WC_PT1 BY Group WITH WC_Pre /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /EMMEANS=TABLES(OVERALL) WITH(WC_Pre=MEAN) /EMMEANS=TABLES(Group) WITH(WC_Pre=MEAN) COMPARE ADJ(LSD) /PRINT=ETASQ HOMOGENEITY DESCRIPTIVE /CRITERIA=ALPHA(.05) /DESIGN=WC_Pre Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Dependent Variable: Word Count Post-test 1 Group Mean Std. Deviation N Group A 98.00 70.529 11 Group B 165.45 134.424 11 Total 131.73 110.296 22 Levene's Test of Equality of Error Variancesa Dependent Variable: Word Count Post-test 1 F 4.344 df1 df2 1 Sig. 20 .050 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WC_Pre + Group ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 106 Tests of Between-Subjects Effects Dependent Variable: Word Count Post-test 1 Type III Sum of Partial Eta Source Squares Corrected Model 118053.712a df Mean Square 2 59026.856 8.162 .003 .462 Intercept 5622.562 1 5622.562 .777 .389 .039 WC_Pre 93028.076 1 93028.076 12.863 .002 .404 Group 25944.511 1 25944.511 3.587 .074 .159 Error 137414.651 19 7232.350 Total 637214.000 22 Corrected Total 255468.364 21 a. R Squared = .462 (Adjusted R Squared = .405) Estimated Marginal Means 1. Grand Mean Dependent Variable: Word Count Post-test 1 95% Confidence Interval Mean 131.727a Std. Error Lower Bound 18.131 Upper Bound 93.778 169.676 a. Covariates appearing in the model are evaluated at the following values: Word Count Pre-test = 150.73. 2. Group Estimates Dependent Variable: Word Count Post-test 1 95% Confidence Interval Group Mean Std. Error Lower Bound Upper Bound Group A 97.385a 25.642 43.715 151.054 Group B 166.070a 25.642 112.400 219.739 a. Covariates appearing in the model are evaluated at the following values: Word Count Pre-test = 150.73. Pairwise Comparisons Dependent Variable: Word Count Post-test 1 F Sig. Squared ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 107 95% Confidence Interval for Differencea Mean Difference (I-J) Std. Error Sig.a (I) Group (J) Group Lower Bound Upper Bound Group A Group B -68.685 36.264 .074 -144.587 7.217 Group B Group A 68.685 36.264 .074 -7.217 144.587 Based on estimated marginal means a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Dependent Variable: Word Count Post-test 1 Partial Eta Sum of Squares Contrast Error df Mean Square 25944.511 1 25944.511 137414.651 19 7232.350 F 3.587 Sig. Squared .074 The F tests the effect of Group. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. UNIANOVA WC_PT2 BY Group WITH WC_Pre /METHOD=SSTYPE(3) /INTERCEPT=INCLUDE /EMMEANS=TABLES(OVERALL) WITH(WC_Pre=MEAN) /EMMEANS=TABLES(Group) WITH(WC_Pre=MEAN) COMPARE ADJ(LSD) /PRINT=ETASQ HOMOGENEITY DESCRIPTIVE /CRITERIA=ALPHA(.05) /DESIGN=WC_Pre Group. Univariate Analysis of Variance Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Dependent Variable: Word Count Post-test 2 Group Mean Std. Deviation N Group A 147.55 65.343 11 Group B 114.18 83.242 11 Total 130.86 74.995 22 .159 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 108 Levene's Test of Equality of Error Variancesa Dependent Variable: Word Count Post-test 2 F df1 2.032 df2 1 Sig. 20 .169 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WC_Pre + Group Tests of Between-Subjects Effects Dependent Variable: Word Count Post-test 2 Type III Sum of Source Squares Partial Eta df Mean Square F Sig. Squared 57383.294a 2 28691.647 8.977 .002 .486 Intercept 19342.643 1 19342.643 6.052 .024 .242 WC_Pre 51261.066 1 51261.066 16.038 .001 .458 Group 5791.131 1 5791.131 1.812 .194 .087 Error 60727.297 19 3196.174 Total 494867.000 22 Corrected Total 118110.591 21 Corrected Model a. R Squared = .486 (Adjusted R Squared = .432) Estimated Marginal Means 1. Grand Mean Dependent Variable: Word Count Post-test 2 95% Confidence Interval Mean 130.864a Std. Error 12.053 Lower Bound Upper Bound 105.636 156.091 a. Covariates appearing in the model are evaluated at the following values: Word Count Pre-test = 150.73. 2. Group ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 109 Estimates Dependent Variable: Word Count Post-test 2 95% Confidence Interval Group Mean Std. Error Lower Bound Upper Bound Group A 147.089a 17.046 111.411 182.767 Group B 114.638a 17.046 78.960 150.317 a. Covariates appearing in the model are evaluated at the following values: Word Count Pre-test = 150.73. Pairwise Comparisons Dependent Variable: Word Count Post-test 2 95% Confidence Interval for Differencea Mean Difference (I-J) Std. Error Sig.a (I) Group (J) Group Lower Bound Upper Bound Group A Group B 32.450 24.108 .194 -18.007 82.908 Group B Group A -32.450 24.108 .194 -82.908 18.007 Based on estimated marginal means a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Dependent Variable: Word Count Post-test 2 Partial Eta Sum of Squares Contrast Error df Mean Square 5791.131 1 5791.131 60727.297 19 3196.174 F 1.812 Sig. Squared .194 The F tests the effect of Group. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. .087 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 110 Appendix J: SPSS Data Analysis for Question 3 GLM WA_PT1 WA_PT2 BY Group WITH WA_Pre /WSFACTOR=time 2 Polynomial /METHOD=SSTYPE(3) /PLOT=PROFILE(time*Group) /EMMEANS=TABLES(Group) WITH(WA_Pre=MEAN)COMPARE ADJ(LSD) /EMMEANS=TABLES(time) WITH(WA_Pre=MEAN)COMPARE ADJ(LSD) /EMMEANS=TABLES(Group*time) WITH(WA_Pre=MEAN) /PRINT=DESCRIPTIVE ETASQ HOMOGENEITY /CRITERIA=ALPHA(.05) /WSDESIGN=time /DESIGN=WA_Pre Group. General Linear Model Within-Subjects Factors Measure: MEASURE_1 Dependent time Variable 1 WA_PT1 2 WA_PT2 Between-Subjects Factors Value Label Group N 0 Group A 11 1 Group B 11 Descriptive Statistics Group Writing Attitude Post-test1 Writing Attitude Post-Test 2 Mean Std. Deviation N Group A 62.09 14.384 11 Group B 59.45 21.667 11 Total 60.77 17.997 22 Group A 59.27 15.363 11 Group B 60.73 20.130 11 Total 60.00 17.490 22 ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 111 Box's Test of Equality of Covariance Matricesa Box's M 5.731 F 1.703 df1 3 df2 72000.000 Sig. .164 Tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups. a. Design: Intercept + WA_Pre + Group Within Subjects Design: time Multivariate Testsa Partial Eta Effect time time * WA_Pre Value Error df Sig. Squared .005 .092b 1.000 19.000 .765 .005 Wilks' Lambda .995 .092b 1.000 19.000 .765 .005 Hotelling's Trace .005 .092b 1.000 19.000 .765 .005 Roy's Largest Root .005 .092b 1.000 19.000 .765 .005 Pillai's Trace .006 .124b 1.000 19.000 .728 .006 .994 .124b 1.000 19.000 .728 .006 Hotelling's Trace .007 .124b 1.000 19.000 .728 .006 Roy's Largest Root .007 .124b 1.000 19.000 .728 .006 Pillai's Trace .025 .479b 1.000 19.000 .497 .025 Wilks' Lambda .975 .479b 1.000 19.000 .497 .025 Hotelling's Trace .025 .479b 1.000 19.000 .497 .025 .025 .479b 1.000 19.000 .497 .025 Roy's Largest Root a. Design: Intercept + WA_Pre + Group Within Subjects Design: time b. Exact statistic Hypothesis df Pillai's Trace Wilks' Lambda time * Group F ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 112 Tests of Within-Subjects Effects Measure: MEASURE_1 Type III Sum Source time time * WA_Pre time * Group Error(time) Partial Eta of Squares df Mean Square F Sig. Squared Sphericity Assumed 8.476 1 8.476 .092 .765 .005 Greenhouse-Geisser 8.476 1.000 8.476 .092 .765 .005 Huynh-Feldt 8.476 1.000 8.476 .092 .765 .005 Lower-bound 8.476 1.000 8.476 .092 .765 .005 Sphericity Assumed 11.479 1 11.479 .124 .728 .006 Greenhouse-Geisser 11.479 1.000 11.479 .124 .728 .006 Huynh-Feldt 11.479 1.000 11.479 .124 .728 .006 Lower-bound 11.479 1.000 11.479 .124 .728 .006 Sphericity Assumed 44.260 1 44.260 .479 .497 .025 Greenhouse-Geisser 44.260 1.000 44.260 .479 .497 .025 Huynh-Feldt 44.260 1.000 44.260 .479 .497 .025 Lower-bound 44.260 1.000 44.260 .479 .497 .025 Sphericity Assumed 1755.430 19 92.391 Greenhouse-Geisser 1755.430 19.000 92.391 Huynh-Feldt 1755.430 19.000 92.391 Lower-bound 1755.430 19.000 92.391 Tests of Within-Subjects Contrasts Measure: MEASURE_1 Type III Sum of Squares Partial Eta Source time df Mean Square time Linear 8.476 1 8.476 .092 .765 .005 time * WA_Pre Linear 11.479 1 11.479 .124 .728 .006 time * Group Linear 44.260 1 44.260 .479 .497 .025 Error(time) Linear 1755.430 19 92.391 Levene's Test of Equality of Error Variancesa F Sig. Squared ASSISTIVE TECHNOLOGY TO ENHANCE WRITING F Writing Attitude Post-test1 Writing Attitude Post-Test 2 df1 df2 113 Sig. .007 1 20 .936 1.052 1 20 .317 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + WA_Pre + Group Within Subjects Design: time Tests of Between-Subjects Effects Measure: MEASURE_1 Transformed Variable: Average Type III Sum of Source Squares Partial Eta df Mean Square F Sig. Squared Intercept 564.244 1 564.244 3.662 .071 .162 WA_Pre 8481.470 1 8481.470 55.044 .000 .743 2.190 1 2.190 .014 .906 .001 2927.621 19 154.085 Group Error Estimated Marginal Means 1. Group Estimates Measure: MEASURE_1 95% Confidence Interval Group Mean Std. Error Lower Bound Upper Bound Group A 60.163a 2.647 54.622 65.704 Group B 60.610a 2.647 55.069 66.151 a. Covariates appearing in the model are evaluated at the following values: Writing Attitude Survey Pre-test = 64.50. Pairwise Comparisons Measure: MEASURE_1 Mean (I) Group (J) Group Difference (I-J) 95% Confidence Interval for Std. Error Sig.a Differencea ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 114 Lower Bound Upper Bound Group A Group B -.446 3.745 .906 -8.286 7.393 Group B Group A .446 3.745 .906 -7.393 8.286 Based on estimated marginal means a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Measure: MEASURE_1 Partial Eta Sum of Squares Contrast Error df Mean Square 1.095 1 1.095 1463.810 19 77.043 F Sig. .014 Squared .906 .001 The F tests the effect of Group. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. 2. time Estimates Measure: MEASURE_1 95% Confidence Interval time Mean 1 60.773a Std. Error Lower Bound Upper Bound 2.297 55.965 65.580 2 60.000a 2.435 54.904 65.096 a. Covariates appearing in the model are evaluated at the following values: Writing Attitude Survey Pre-test = 64.50. Pairwise Comparisons Measure: MEASURE_1 95% Confidence Interval for Differencea Mean Difference (I-J) Std. Error Sig.a (I) time (J) time Lower Bound Upper Bound 1 2 .773 2.898 .793 -5.293 6.839 2 1 -.773 2.898 .793 -6.839 5.293 Based on estimated marginal means a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). ASSISTIVE TECHNOLOGY TO ENHANCE WRITING 115 Multivariate Tests Partial Eta Value F Hypothesis df Error df Sig. Squared Pillai's trace .004 .071a 1.000 19.000 .793 .004 Wilks' lambda .996 .071a 1.000 19.000 .793 .004 Hotelling's trace .004 .071a 1.000 19.000 .793 .004 Roy's largest root .004 .071a 1.000 19.000 .793 .004 Each F tests the multivariate effect of time. These tests are based on the linearly independent pairwise comparisons among the estimated marginal means. a. Exact statistic 3. Group * time Measure: MEASURE_1 95% Confidence Interval Group time Mean Group A 1 61.553a 3.249 54.752 68.354 2 58.773a 3.444 51.564 65.982 1 59.992a 3.249 53.191 66.793 2 61.227a 3.444 54.018 68.436 Group B Std. Error Lower Bound Upper Bound a. Covariates appearing in the model are evaluated at the following values: Writing Attitude Survey Pre-test = 64.50.