Writing has been called the "neglected R'" in the traditional trilogy of reading, riting, and `rithmetic (National Commission on Writing, 2003). Writing performance continues to languish, despite societal expectations that students should be able to write clearly and precisely. Sociocognitive theory predicts that writing beliefs are related to writing performance. Much research has focused on writing self-efficacy beliefs and their link to writing apprehension and writing performance, while research exploring another type of belief, domain-specific beliefs about writing itself, is sparse. This study examined the relations between these beliefs about writing, writing self-efficacy, and writing apprehension, and their links to writing performance.

This research was a three-phase study. Phases I and II involved instrument construction and validation, while Phase III examined the relations among the research variables. Two hundred eighty-seven Hispanic women students completed a test battery in class measuring demographics, beliefs about writing, writing self-efficacy, and writing apprehension. Writing performance was measured separately on an authentic writing task, a take-home paper, by both an overall grade and six component grades. Inter-rater agreements on these grades ranged from r = .83 to .91. 

Hierarchical regression analyses revealed that beliefs about writing independently predicted writing performance and that some beliefs about writing (e.g., Good writers adapt their message to their readers) are adaptive and associated with strong writing performance, while other beliefs about writing (e.g., Readers are impressed by big words) are maladaptive and relate to weak writing performance. In addition, apprehension about making grammatical and other mechanical errors had a stronger negative effect on writing performance than the more traditional concept of writing apprehension, which concerns sharing one's writing with others and having it critiqued. After controlling for domain-specific beliefs, writing self-efficacy weakly predicted writing performance as well. 

These results support the need for future research examining the relations among the research variables and writing performance in samples that are more balanced with respect to gender and ethnicity, and with other writing tasks. Because beliefs about writing demonstrated the largest beta weights in the regression equations, these beliefs may have the most promise for promoting both writing research and practice.