Using Bayesian statistics to investigate the relationships among l2 writing attitude, writing self-efficacy and writing anxiety

The role of relevant psychological variables in L2 writing is very well documented in the literature and writing attitude, writing self-efficacy and writing anxiety are no exception. However, it is seen in the literature that the majority of the studies with respect to L2 writing-related psychological variables utilize frequentist or classical statistics to make inferences and come to conclusions. In this respect, the present study aimed to fill a gap in the literature by proposing an alternative interpretation of the possible relationships among L2 writing attitude, writing self-efficacy and writing anxiety by means of Bayesian statistics, as they have advantages over their frequentist counterparts in educational research in terms of estimating more complex models, making estimations based on prior models and being more informative in general. The study had a cross-sectional and descriptive design. 152 pre- service teachers of English participated in the study and completed writing attitude, writing self-efficacy and writing anxiety scales voluntarily. The data collected was subjected to analyses to check for normality. Following the tests of normality, Bayesian correlation analyses were used to check potential relationships among the aforementioned psychological variables. The findings confirmed the previously identified positive correlation between L2 writing attitude and writing self-efficacy as well as the negative correlation of L2 writing anxiety with writing attitude and writing self-efficacy with a Bayesian approach, indicating extreme evidence for all relationships