Moving Beyond Significance: Understanding and Reporting Effect Sizes in Applied Linguistics Research
VI. International Applied Statistics Congress (UYIK – 2025) , Ankara, Türkiye, 14 - 16 Mayıs 2025, ss.159, (Özet Bildiri)
- Yayın Türü: Bildiri / Özet Bildiri
- Basıldığı Şehir: Ankara
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.159
- Karadeniz Teknik Üniversitesi Adresli: Evet
Özet
In applied linguistics research, the overreliance on p-values as the primary indicator of "significant" results often overshadows the importance of effect size, which provides a measure of the magnitude of the difference between groups or association between variables. Recent methodological critiques within the social sciences—and increasingly in applied linguistics research—have called for a shift beyond simplistic, dichotomous thinking around statistical outcomes, a movement referred to as a 'methodological turn’ and characterized by heightened 'methodological awareness'. This presentation critically examines the limitations of statistical significance as the primary method for determining the meaningfulness of results and advocates for the inclusion of alternative or complementary approaches, such as effect size interpretation, confidence intervals, and the consideration of practical significance. Reporting effect sizes alongside p-values enhances both the transparency and practical relevance of research findings. Recognizing the practical importance of effect size enhances the interpretability of findings and enables meaningful comparisons across studies in applied linguistics. This presentation advocates for an approach to statistical analysis that balances methodological precision with practical relevance, urging applied linguists to look beyond p-values by incorporating measures of effect size and confidence intervals, reported alongside significance testing, to yield more meaningful interpretations of research findings.