An Assessment of the Writing Skills and Writing Self-Efficacy of Autistic College Students and Their Mentors

Background:

Discrepancies between high cognitive and variable writing skills have been documented since the earliest accounts of autism (Asperger, 1991). Although writing skills are essential in college, the writing skills and self-efficacy of autistic college students have received almost no empirical attention (Gerstle & Walsh, 2010). Jurecic (2007) described difficulties one autistic college student faced writing for an audience and speculated that they might be attributable to Theory of Mind (ToM) challenges. The few studies that have examined writing skills among autistic children/adolescents suggest that they produce fewer written words and have more variable writing skills but do not differ overall in standardized writing scores relative to non-autistic youth (Griswold et al., 2002; Myles et al., 2003).

Objectives:

1) To compare writing skills, self-efficacy, and ToM among autistic and non-autistic college students.

2) To examine associations between measures among autistic students.

Methods:

Autistic college students in a mentorship program (n = 19) and non-autistic mentors (n = 6) completed pre-test assessments: the Social Responsiveness Scale-2 (SRS-2); Woodcock-Johnson Word Comprehension; Reading the Mind in the Eyes (RMIE), a nonverbal ToM measure; a writing self-efficacy measure (MacArthur et al., 2016); the Test of Nonverbal Intelligence (TONI); and a writing activity, wherein they were asked to share something interesting they had learned recently.

Results:

Autistic students reported heightened autistic traits and belief in writing conventions (a subscale of the self-efficacy scale; e.g., “Good writers don’t make errors in grammar”), wrote fewer words and sentences, and had lower RMIE scores (ps< .05) relative to mentors (see Table 1). Word comprehension, NVIQ, nonverbal ToM, other subscales of the self-efficacy measure (self-efficacy, goal-oriented avoidance, performance and mastery, beliefs about content and writing affect), errors, examples and perspective taking in writing did not differ.

Among autistic students, higher SRS-2 scores were positively associated with word comprehension and negatively correlated with goal-oriented performance (ps< .03). Nonverbal intelligence was associated with nonverbal ToM; word comprehension was associated with RMIE (ps< .02). No associations between writing volume, writing self-efficacy subscales, or ToM measures were observed.

Conclusions:

Replicating findings with autistic children/adolescents, autistic college students produced shorter texts than non-autistic mentors despite no group differences in NVIQ and word comprehension. This is consistent with our observations during mentorship; a number of autistic students encounter pronounced difficulty initiating and elaborating on their ideas through writing despite often high grammatical skills. Heightened belief in writing conventions (perfectionistic/rigid beliefs about writing) may contribute to these difficulties. However, no associations between writing self-efficacy subscales and writing performance were observed. The new writing self-efficacy measure used in this study had not previously been used with autistic students; it was highly reliable in this sample (α= .92). Autistic students’ provided heterogeneous accounts of their writing self-efficacy, warranting further examinations of predictors of writing self-efficacy in a broader sample. Consistent with evidence that associations between ToM and spoken narrative skills among autistic children are apparent only for children with below average IQs (Capps et al., 2000; Losh & Capps, 2003), ToM was unrelated to writing skills among autistic students with average IQs.