Top Five Statistical Methods Used in Applied Linguistics: Focusing on (hierarchical) regression

Link to the Zoom meeting:

https://hkbu.zoom.us/j/91565666136 Meeting ID: 915 6566 6136

This talk presents an overview of the five most commonly used statistical methods: t-test, ANOVA, correlation, chi-square test, and regression (Wei et al., 2019); among these “Top Five” methods, regression is arguably most useful (Plonsky & Oswald, 2017). This talk highlights the importance of effect size with examples from published research because effect size is deemed more important than the statistical significance level (viz. the p value), which tends to be the only focus in the “old statistics” paradigm (cf. Cumming, 2014). This talk also illustrates the benefits and pitfalls concerning the use of hierarchical regression, which is most instrumental in ascertaining the unique contribution of each predictor (independent variable) to the dependent variable, as this variant of regression is precited to “be used more frequently” (Wei et al., 2020:7) in future applied linguistic studies.

Rining WEI (Tony), PhD, is a tenured Associate Professor in Xi’an Jiaotong-Liverpool University (XJTLU), China’s largest Sino-foreign institution. He has supervised nearly 40 master’s and PhD dissertations covering topics such as English for Academic Purposes (EAP) and individual differences in L2 learning. His areas of research include multilingualism, geronto-linguistics, and quantitative methodology. He has published in journals including Bilingualism: Language & Cognition, English Today, Journal of Multilingual and Multicultural Development, Journalism, Linguistics and Education, System, and World Englishes. He serves as (1) Associate Editor for TESOL International Journal (Scopus-indexed), and (2) Co-Editor-in-Chief for International Journal of EAP: Research and Practice.