A practical guide to bootstrapping descriptive statistics, correlations, T tests, and anovas

Geoffrey T. LaFlair, Jesse Egbert, Luke Plonsky

Research output: Chapter in Book/Report/Conference proceedingChapter

18 Scopus citations

Abstract

This chapter provides a practical and detailed account of how to run, interpret, and report bootstrapped analyses. Because of their frequency of use in the field, that focuses on four types of analyses/statistics: descriptives, t tests, ANOVAs, and correlations. These bootstrapped analyses are accompanied with confidence interval(CIs) for the statistic of interest. The chapter also explains bootstrapping that has the potential to be a powerful nonparametric analytical tool when L2 researchers faces with problems such as small samples and nonnormal distributions. It reports bias corrected and accelerated(BCa) intervals from both SPSS and the boot package in R. In the SPSS examples, all bootstrapped analyses have been performed using the simple resampling method. The procedure jackknife-after-boot, useful for investigating the effect of outliers on the bootstrapped calculations. This chapter also focuses for a diminished role of the flawed and unreliable practice of statistical significance testing and instead for a greater emphasis on descriptive statistics-namely means, standard deviations, CIs, and effect sizes.

Original languageEnglish (US)
Title of host publicationAdvancing Quantitative Methods in Second Language Research
PublisherTaylor and Francis
Pages46-77
Number of pages32
ISBN (Electronic)9781317974093
ISBN (Print)9780415718332
DOIs
StatePublished - Jan 1 2015

ASJC Scopus subject areas

  • Arts and Humanities(all)
  • Social Sciences(all)

Fingerprint

Dive into the research topics of 'A practical guide to bootstrapping descriptive statistics, correlations, T tests, and anovas'. Together they form a unique fingerprint.

Cite this