Statistical power, P values, descriptive statistics, and effect sizes: A “back-to-basics” approach to advancing quantitative methods in L2 research

Luke Plonsky

Research output: Chapter in Book/Report/Conference proceedingChapter

67 Scopus citations

Abstract

This chapter reviews major weaknesses of relying on statistical significance p values, particularly in the case of tests comparing means such as t tests, ANOVAs and correlations. Along with a brief introduction to the notion of statistical power, followed by guides to calculating and using effect sizes and other descriptive statistics including confidence intervals. A related descriptive statistic that is considered and reported even less frequently is the CI. As an alternative, the thorough use of descriptive statistics, including effect sizes and CIs, should replace much of the statistical testing in L2 research. The main reasons for using effect sizes largely correspond to and address the flaws of NHST. However, primary researchers currently do little in the way of using effect sizes to enhance the results or, more importantly, the understanding of the variables and relationships. Both statistical and practical significance are considered and interpreted, and the results of the study are brought together via research synthesis and meta-analysis.

Original languageEnglish (US)
Title of host publicationAdvancing Quantitative Methods in Second Language Research
PublisherTaylor and Francis
Pages23-46
Number of pages24
ISBN (Electronic)9781317974093
ISBN (Print)9780415718332
DOIs
StatePublished - Jan 1 2015

ASJC Scopus subject areas

  • General Arts and Humanities
  • General Social Sciences

Fingerprint

Dive into the research topics of 'Statistical power, P values, descriptive statistics, and effect sizes: A “back-to-basics” approach to advancing quantitative methods in L2 research'. Together they form a unique fingerprint.

Cite this