Improving Writing Skills in Children with ADHD

Writing is a challenging task that requires attention to multiple processes including planning, generating content, organization one's ideas, translating ideas into written language, and editing/improving the writing. Skillful writers are adept at using specific strategies to help them plan, organize, write, and revise their compositions, as well in using strategies to monitor the quality of their efforts.

Given the multiple demands required for good writing, it not surprising that children with ADHD struggle with this academic task. In addition to well documented difficulties with neatness/penmanship, as well as with spelling, there is emerging evidence that the general composition skills in children with ADHD lags behind their peers.  Specifically, children with ADHD tend to produce stories that are less complex, less coherent, and significantly shorter.  These deficits in written expression may reflect difficulties with important executive functions such as planning, organization, and working memory skills that characterize many children with ADHD.

A paper published in a recent issue of Exceptional Children (Reid, R., & Lienemann, T.O., 2006. Self-regulated strategy for development of written expression with students with Attention Deficit Hyperactivity Disorder, Exceptional Children, 73, 53-67) presents encouraging preliminary data on the use of a systematic approach to improve writing skills in children with ADHD.  Participants in this study were 3 9-10 year-old children - 2 boys and 1 girl - who were diagnosed with ADHD and who scored in the bottom 20% on a standardized test of written language.  All 3 children were receiving medication treatment at the time of the study.


Self-Regulated Strategy Development Instructional Approach

Children were provided direct and systematic instruction in procedures for planning and writing coherent, complete, and well-organized stories in 4 one-on-one lessons.  The lessons included specific strategies for writing stories as well as instruction in monitoring one's writing.  These lessons are summarized below and can also be viewed in its entirety at http://kc.vanderbilt.edu/casl/srsd.html


Lesson 1 - Develop Background Knowledge

This lesson used mnemonics to teach students about the process of story writing, and the important parts of a good story.  The first mnemonic introduced was "POW" which stands for:

* Pick my ideas
* Organize my notes
* Write and say more

Students were asked to explain these steps to the instructor and the student and instructor discussed the importance of using POW when writing a story.

A second mnemonic, WWW, What=2 and How=2, was introduced to teach students that good stories include a number of key elements that answered important questions for the reader. These were:

"Who are the main characters?"
"Where does the story take place?"
"When does the story take place?"
"What do the main characters want to do?"
"What happens next?"
"How does the story end?"
"How does the main character feel?"

After reviewing these elements, the instructor read students a story and had them pick out the different parts.  As students identified each element, the instructor wrote their responses on a chart to provide a pictorial representation to help them remember the parts of a good story.  This process was repeated with a second story and students were reminded that a "good story" included all these parts.  Students were introduced to the idea of looking for these parts in their own stories as a way to monitor their writing.

Before leaving, students were instructed to memorize POW and WWW, What=2 and How=2 before the next lesson.


Lesson 2 - Discuss It

This lesson provided continued practice and review of the strategies introduced in lesson 1.  First, students practiced finding the important story parts in stories that were read to them.  Next, they analyzed a story they had written prior to beginning instruction to see which story elements they included. Students were taught to graph the number of parts included as well as the number of words; the goal was to actively involve students in monitoring their progress writing. The instructor and students discussed that the "goal" in writing a good story was to make sure that it "made sense", used "million-dollar words", was fun to write, and included all seven parts.


Lesson 3 - Model It

In this lesson, the instructor modeled the process of using POW and WWW, What=2 and How=2.  Before beginning a story, the instructor discussed the goals in writing a good story, and, with help from the student, generated a story in response to a picture prompt that included all 7 parts.  As ideas were generated, the instructor recorded them on a planning sheet, which included a prompt for the seven parts. 

The instructor modeled the use of self-statements during the writing such as "What is my goal?", "What is my next step?", and "Is this story making sense?". The importance of using "self-talk" to help guide and monitor one's writing was discussed and students were asked to create 3 self-statements they could use to help their writing. 

After the story was finished, the students and instructor checked to make sure it included all 7 parts; they also graphed the parts and the number of words, with the instructor noting that longer stories could often be better stories.


Lesson 4 - Support It

In this lesson, students wrote a story with the instructor using the  strategies taught in earlier lessons.  The students directed the writing with the instructor providing support as needed.  After completing the story, the student checked for the 7 key story elements and the student and instructor discussed how using the strategies would help the student write better stories.


Did this instruction approach improve students writing?

To evaluate the impact of the writing program, the researchers used a "multiple baseline design", a common method for studies with a small number of subjects.  At the start of the study, each child completed at least 3 stories on their own.  These "baseline" stories were written in response to picture prompts, i.e., students were asked write a "good story" about pictures that were shown to them using as much time needed.  The stories written by each child were evaluated in terms of:

- the number of story elements included;
- the number of words used in the story;
- an overall rating of the story quality made on a 1 to 7 scale;

The instructional process began for each child after varying time intervals, such that the number of baseline sessions, i.e., stories written prior to instruction, was longer for the 2nd child than the 1st and for the 3rd child than the 2nd. 

Immediately after each child completed the 4-session training program, they independently wrote 3 additional stories; this enabled the researchers to determine whether the quality of their stories had changed after instruction. 

Because the length of time between the baseline stories and post-training stories varied for each child, the researchers could tell whether any improvement in children's writing that was evident was attributable to the instruction itself, or merely to the passage of time.  The former would be the case if children's writing showed improved only after they had received instruction, regardless of how long their "baseline" period was.  The latter would be the case if children with longer baselines, i.e., those for whom the start of training was delayed, showed improvement prior to when instruction began.  This could occur if the regular instruction they were receiving in the classroom was helping them to write better.

Because instruction began at different times for the 3 children, the amount of follow-up data necessarily varied.  Thus, for the child beginning first, follow-up was conducted both 3 and 6 weeks after the instruction to see whether gains that were evident persisted.  For the second child, only 3-week follow-up data was available because the school year ended.  Follow-up assessment was not conducted with the 3rd child, who ended instruction at the end of the school year.


Results

- Number of story elements -

During the baseline phase, no student included all 7 parts in any story and the average number of story parts included was fewer than 4 for each child. In addition, the number of story elements included in the different baseline stories was highly variable from story to story. Such inconsistent performance is quite common in children with ADHD.

Immediately following instruction, performance both increased immediately and stabilized.  In fact, each child included all 7 key story elements in all their stories.  For the student who began instruction first, stories written 3- and 6-weeks after instruction continued to include all elements. Improvement was also maintained across 3 weeks for the second child.


- Number of words -

Stories completed after instruction showed a substantial increase in the number of words.  The average number of words contained in the baseline stories for the 3 children was 33, 29, and 42; after instruction, this increased to 225, 60, and 169 words, increases of 681%, 206%, and 323% respectively.  For the 2 children on whom follow-up data was collected, some decline in story length was evident but story length still remained significantly above baseline.


- Holistic quality -

It is possible that although children were writing longer stories that included more important elements, they were still not well-organized and were difficult to follow.  However, the holistic rating of story quality indicated that this was not the case.  Prior to instruction, the average holistic rating for the children's stories was 1.3, 1.5, and 1.8; thus, each child's stories were rated towards the bottom of the 1 to 7 scale of overall quality.

After treatment, holistic ratings increased to 5.3, 2.8, and 4.7 respectively.  For the 2 children on whom follow-up data was available, this improvement in overall story quality of persisted. 

It is important to note that quality ratings were made by raters who did not know whether the stories were written during baseline, immediately following instruction, or during the follow-up periods. Thus, these ratings were not biased by knowledge of when they had been written.


Summary and Implications

Results from this study indicate that the strategy instruction approach used can improve narrative writing skills in children with ADHD. This improvement was evident in the inclusion of key story elements, story length, and also in the overall quality and coherence of children's stories.  Use of a multiple baseline design indicates that these improvements were linked to the strategy instruction, and not to the passage of time and additional exposure to regular classroom writing instruction.

In addition to these pronounced effects on writing quality, the decrease in variability of children's stories is also noteworthy. Although the stories following instruction were collected across different days, they were consistently higher quality and showed less of the day-to-day variability that is commonly seen in students with ADHD. 

The authors suggest that this reduction in variability occurred in response d to the "self-monitoring" aspects of the intervention.  That is, they believe that teaching children to use "self-talk" during the writing task, and to review and graph stories for the key elements enabled children to perform more consistently.  As they note, however, the study design does not enable them to determine which component of the intervention package - the specific writing strategy instruction, the self-monitoring of performance, or both - was responsible for this.  Teasing this apart would require a larger study in which these different intervention components were employed both separately and in unison.


- Limitations -

Although the results of this study are encouraging, the authors point out a number of limitations that are important to recognize.  First, the sample size was obviously small.  Replicating these findings with additional "single subject" designs, as well as in larger group studies, would thus be essential for documenting the value of the writing instruction approach used here.

Second, the follow-up period was quite brief - 6 weeks for 1 child and only 3 weeks for a second.  The impact of the instruction program on children's writing over a sustained time period is thus unknown.  Because long-term maintenance of academic improvement following intervention is a serious concern for students with ADHD, determining whether these gains persist over a longer period is essential.

Third, all instruction was done in a one-on-one setting which may be difficult to replicate in regular school settings.  Determining whether this instruction can be effectively implemented in a small-group setting would thus be important to pursue.

Finally, the children in this study were all on medication during the instruction and across all data collection periods.  Whether this approach would work equally well with unmedicated students remains to be determined.

Despite these limitations, which should be addressed in subsequent research, these results are encouraging in that they suggest that a sound and systematic instructional approach can help elementary school students with ADHD make important gains in the quality of their writing.  I hope that the authors will address the limitations noted above in their subsequent work, and that the promise shown in this preliminary study will be fulfilled.