ABSTRACT
Causal-comparative research, like experimental research, studies group differences, but these differences already exist when the study is conducted. Such group differences may include gender, L1, field of study, school district, or other demographic variables that cannot be manipulated and that require labeling participants according to nominal variables. Cheng (2002), for example, examined gender differences in L2 writing anxiety, which was measured through the use of a questionnaire, and found that females had higher L2 writing anxiety. In addition, he looked at the relationship between language anxiety and language class level. In this case, class level was treated as a nominal variable (i.e., placement into one of three groups based on year in school), and an ANOVA was used to test for group differences. In contrast, if placement test scores (a continuous variable) had been used instead of class placement, this part of the study would have been correlational (discussed later). In neither case, however, can a causeeffect relationship between class level and anxiety be assumed; language anxiety can negatively affect learning and poor progress can affect anxiety. Doolan and Miller (2012), another causal-comparative study, examined differences in the errors of L1 and Generation 1.5 writers, that is, writers distinctly different from international students in that they have lived in the US for an extended period of time and usually have strong oral skills. The authors found that Generation 1.5 writers made more errors overall and more of certain error types. In this case, we do not know what factors (e.g., language input at home, educational background, motivation) related to being a Generation 1.5 writer caused the difference in errors.
