The Effects of a Motivational Intervention on Improving the Writing Productivity of Students with Learning Disabilities

The purpose of this study was to implement and evaluate a motivational intervention that consisted of explicit timing, immediate feedback through self-scoring, and display of high scores with the goal of encouraging four students with learning disabilities to write longer stories. All participants were initially very unwilling to engage in text production. An ABA reversal design was used to assess the effectiveness of the approach. Immediately after the treatment, the subjects started to write more extensive stories. However, this effect ended abruptly once the intervention was terminated. The study’s results show that, even for very reluctant students with learning disabilities, writing motivation can be significantly enhanced with relatively little effort. All the participants reported enjoying the treatment. The paper ends with a discussion of the experiment’s limitations and the practical implications of the findings.

However, even the most effective approach will not yield the intended results, if students refuse to write because they deem it to be too laborious and wearisome. In a widely cited article on the quality indicators for effective text-production training entitled "What is missing from current writing intervention programs?", De Caso and García (2006) suggested a solution: writing motivation programs. Helpfully, many time-tested educational arrangements focus on influencing the processes that initiate, guide, and maintain students' goal-oriented behaviors.
In several studies, researchers have successfully implemented the strategies of explicit timing, providing immediate feedback through self-scoring, and displaying high scores to increase the motivation to produce texts among children with learning disabilities (Van Houten, Hill, & Parsons, 1976;Van Houten & McKillop, 1977;Van Houten, Morrison, Jarvis, & McDonald, 1974). When applying explicit timing, the teacher gives the students a certain task, informs them of the time interval to complete it, and observes them as they try to be as effective as possible within the given time frame (Van Houten & Thompson, 1976). When immediate feedback is provided through self-scoring, learners monitor their own performance by assessing the relevant target variable at the end of an assignment (Light, McKeachie, & Lin, 1988). Prominently displaying students' high scores (e.g., on a poster hung on the classroom wall) is an effective positive reinforcement technique that has often been used in school settings (Archer & Hughes, 2011;Prater, 2018).
There are several fairly recent studies on the effectiveness of the aforementioned motivational methods. For example, Duhon, House, Hastings, Poncy, and Solomon (2015), Grays, Rhymer, and Swartzmiller (2017) as well as Wells, Sheehey, and Sheehey (2017) used these techniques to improve mathematics fluency in elementary and middle school students. McDaniel, Jolivette, and Ennis (2013) demonstrated their positive impact the on the oral reading ability of a third grader with emotional and behavioral problems. But even though research interest in motivational systems to increase academic productivity is still alive in the scientific community, there are only two published papers on the aforementioned methods for improving writing performance in students with learning disabilities: Grünke, Sperling, and Burke (2017) and Grünke, Knaak, and Hisgen (2018). In both studies, explicit timing, immediate feedback through self-scoring, and displayed high scores were extremely effective; the children (who were 10 or 11 years old) were initially very unmotivated but went on to produce comparatively long texts. The purpose of this case report was to add to the limited existing findings on the benefits of these motivational writing-productivity techniques by considering a sample of four students with learning disabilities who were slightly older than those in the two prior experiments.

Setting and Participants
The four participants are labeled here as Adrian, Bella, Christin, and Dominik (names changed for anonymity). Bella and Christin (both female) were 14 years old, and Adrian and Dominik (both male) were 12 years old. All four attended the same special school for students with moderate general-learning disabilities. It is located in a rural area outside of a major city in North Rhine-Westphalia (Germany). None of the students came from an immigrant background. All had been diagnosed with a learning disability by a multi-professional team and demonstrated a distinctive discrepancy between expected and actual achievement in reading and arithmetic. According to their teacher, their spelling skills were acceptable, but they were not able to compose text well. Whenever required to write something, they usually produced only one or two short sentences.

Dependent Variable and Experimental Design
Writing performance was assessed using a general-outcome measure of overall writing ability (McMaster & Espin, 2007) consisting of counting the number of total words written (TWW), disregarding spelling errors (Furey, Marcotte, Hintze, & Shackett, 2016). This measurement's reliability was determined by having two observers independently review the texts: an interventionist (see below) and a student assistant. The interrater agreement was 100%.
An ABA reversal plan (Ledford & Gast, 2018) was employed to analyze the motivational system's effects. This design consisted of a baseline condition (A1), followed by a treatment phase (B) and then a return to the baseline condition (A2). Each of the three phases was composed of five daily probes. Adrian was ill on the first day of the study; apart from that, all four students were present during the whole experiment. The ABA design was appropriate because we expected that TWW would change depending on whether the motivational system was in place. The performance was not anticipated to remain at a high level after the treatment was withdrawn. According to Tawney and Gast (1984), the ABA plan is the simplest single-case design that can identify functional ies.ccsenet. relationshi

Proced
During the (who had room of th phase (A1 produced a them. The motivated narratives At the beg that depict still room index card Figure 1).

Results
The motiv and figure org vational progra s were generat The participants had rather stable baselines, followed by improved performance during the B phase and a return to low scores in the A2 phase. The scores for the A1 and A2 phases, as compared to those in the B phase, showed 80% nonoverlapping data (PND, Scruggs & Mastropieri, 1998) for Adrian and Bella, and 100% nonoverlapping data for Christin and Dominik. In all cases, 100% of the data exceeded the median (PEM, Ma, 2006) and the median trend (PEM-T, Wolery, Busick, Reichow, & Barton, 2010). A piecewise regression analyses (Huitema & McKean, 2000) for the four students yielded the results shown in Table 1. As indicated in Table 1, there were no developmental trends in phase A1. In addition, there were no noteworthy level effects from phase A1 to phase B. However, the steepness of the regression lines differed significantly in all cases except for that of Christin (from A1 to B and from B to A2). The first baseline phase could be considered stable, and there was a continuous treatment effect for Adrian, Bella, and Domenik in the B phase. These three participants continued to improve their performance as the motivational intervention proceeded. As soon as the college student stopped applying explicit timing, requiring self-scoring, and displaying the high scores, the participants' achievements ceased. In fact, all participants (including Christin) demonstrated a sudden decline in their output upon the termination of the treatment, as represented in significant level effects between phases B and A2.
To complete the statistical data analysis, the four cases were aggregated into one using hierarchical linear modeling (see Table 2). Overall, the motivational system seemed to elicit a significant continuous treatment effect, which immediately vanished as soon as the intervention was over.  All students reported that they enjoyed the challenge of trying to beat their previous high scores. They indicated that they usually did not like to write but that the use of the timer and the visibility of their improvements made it much easier for them to attend to the task. They regretted that the college student stopped the motivational system after only one week and stated that they hoped to participate in a project such as this again.

Discussion
The results of this study suggest that a motivational system (consisting of explicit timing, self-scoring, and displaying high scores) has the potential to help students with learning disabilities to write considerably longer stories than they would otherwise. All four participants in this research wrote longer texts in the last trial of the B phase than in any previous trial during the experiment. Every effect-size index (PND, PEM, PEM-T) indicates a very beneficial treatment outcome, with scores between 80 and 100%. Even though the slope effect did not reach statistical significance for Christin, a visual inspection of her performance curve in the B phase still suggests that she benefited from the intervention. In addition, the subjects gave very favorable feedback on the treatment during the informal interviews at the end of the study.
Despite these positive effects, a number limitations of this research should be acknowledged to put the findings in context. First, as only four participants were included, involving more individuals would have strengthened the results and made them more generalizable to the population that Adrian, Bella, Christin, and Dominik were drawn from. Second, the phases were all rather brief. Thus, it is possible that the motivational system would have lost its potency if it had been implemented for longer. The results provide no way of telling how the increase in word production would have developed had the intervention continued for another week or two. It is certainly conceivable that the students would have lost interest in writing stories. A third limitation pertains to the fact that the motivational system was evaluated as a whole, which means that there is no way to determine the isolated benefits of any of the three treatment elements. Fourth, the ABA design does not end on a high note, as the potentially effective intervention is withdrawn in the last step. Subsequently, the participants are left to their own resources and do not receive any further help (Ledford & Gast, 2018). The final critical issue relates to the way in which the dependent variable was measured. Although TWW is certainly a very common index, it is not the only option for capturing writing performance. Future studies should utilize not only TWW, but also other measureslike the number of correct word sequences (CWS) or different writing rubrics (see, e.g., Gansle, Noell, VanDerHeyden, Naquin, & Slider, 2002) -to determine the effectiveness of writing interventions like the one described in this paper. In particular, tools that focus on more than just productivity could be of unique interest. Beyond that, it could be helpful to incorporate such scoring methods into the intervention by employing them to give students continuous feedback on the level of appropriate word choice, the organization of the produced texts, and other indicators. Providing struggling writers frequently with information on how well they are doing might elicit a further increase in performance and help them sustain the developed skills (see e.g. McKeown, Kimball, & Ledford, 2015).
Despite its limitations, this study represents additional empirical support for the use of explicit timing, self-scoring, and displaying high scores for overcoming resistance to writing among students with learning disabilities. Finding ways to encourage these children and adolescents to complete a task that they usually detest is certainly a great challenge. This experiment, however, provides reason for optimism, as its results indicate that it does not require much effort to provide these students with motives to engage in text production. To create proficient writers, it is certainly not enough to encourage learners to compose simple stories. That is only the start. Custom interventions based on the aforementioned SRSD model (or on other sound approaches) need to apply scaffolding to help struggling students on their paths to becoming skillful at producing not only plain narratives but also more elaborate texts.
In conclusion, the advantage of this study is its potential to help learners accomplish a crucial subgoal (engaging in ies.ccsenet.org International Education Studies Vol. 12, No. 3; text composition) with rather little effort. The expense of the intervention was minimal. Most classrooms have timers. Plastic folders and notebook paper are inexpensive. Each session took only few minutes and thus was very efficient. It remains to be seen whether the motivational system evaluated in this study will receive a wide distribution in everyday writing education for students with learning disabilities (or other challenged students). Researchers thus need to address the questions that this experiment has left unanswered.