Statistical modeling of SLA theories: connecting test performance to construct

Yo In'nami, Akira Murakami

Research output: Chapter in Book/Report/Conference proceedingChapter (peer-reviewed)peer-review

Abstract

After discussing the advantages of statistical modeling over the traditional conceptualisation of analytical techniques, this chapter describes two currently underexploited statistical modeling approaches to examining longitudinal second-language development. First, mixed-effects models allow researchers to investigate linear and nonlinear changes while incorporating individual variations and other random effects. Second, latent variable models can be used to investigate linear and nonlinear changes while controlling for measurement error. This chapter discusses how these two approaches can be applied to examine important theoretical issues in second-language development research and also explains some of the techniques and concepts useful for modeling longitudinal data.
Original languageEnglish
Title of host publicationThe Routledge handbook of second language acquisition and language testing
EditorsPaula Winke, Tineke Brunfaut
PublisherRoutledge
Chapter42
Pages445-456
Number of pages12
Edition1st
ISBN (Electronic)9781351034784
ISBN (Print)9781138490680
DOIs
Publication statusPublished - 28 Dec 2020

Publication series

NameThe Routledge Handbooks in Second Language Acquisition
PublisherRoutledge

Fingerprint

Dive into the research topics of 'Statistical modeling of SLA theories: connecting test performance to construct'. Together they form a unique fingerprint.

Cite this