Statistical Literacy Among Applied Linguists and Second Language Acquisition Researchers

Abstract

The importance of statistical knowledge in applied linguistics and second language acquisition (SLA) research has been emphasized in recent publications. However, the last investigation of the statistical literacy of applied linguists occurred more than 25 years ago (Lazaraton, Riggenbach, & Ediger, 1987). The current study undertook a partial replication of this older work by investigating (a) applied linguists' general experiences with statistics, (b) underlying factors that constitute applied linguists' knowledge about and attitudes toward statistics, and (c) variables that predict attitudes toward statistics and statistical self-efficacy. Three hundred thirty-one scholars of applied linguistics and SLA completed a questionnaire. Eighty percent had taken a statistics class; however, only 14% of doctoral students and 30% of professors felt that their statistical training was adequate. A factor analysis of participants' knowledge of statistical terms revealed three factors: common inferential statistics knowledge, advanced statistics knowledge, and basic descriptive statistics knowledge. An analysis of participants' attitudes toward statistics revealed two factors: statistics are important and lack of statistical confidence. Regression analyses found that a quantitative research orientation was the strongest predictor of positive attitudes toward statistics; nevertheless, participants also expressed support for qualitative research. Recommendations for improving quantitative methods in our field are made based on our findings.

Multiple Regression in L2 Research: A Methodological Synthesis and Guide to Interpreting R2 Values

“…Although it is not particular to MR, estimates of instrument reliability (internal consistency) were also found inconsistently in our sample, a finding observed within and across numerous other domains of L2 research (e.g., Plonsky & Derrick, ; Plonsky & Gass, ; Plonsky & Kim, ). The software used to run the MRs in our sample was reported only 12% of the time; as found elsewhere, SPSS was by far found to be used most often (see Loewen et al., ).…”

“…We also hope to contribute to the methodological reform movement currently taking hold in applied linguistics. In addition to the methodological syntheses described previously, evidence for the movement that Byrnes's () editorial referred to as the “methodological turn” (p. 825) can be seen in a variety of activities such as recently revised journal submission guidelines (e.g., Norris et al., ), newly introduced analytical techniques (e.g., Plonsky, Egbert, & LaFlair, ), and a number of recent examinations of statistical literacy (e.g., Loewen et al., ) (for an overview of other activities associated with methodological reform in applied linguistics, see Marsden & Plonsky, ). As a contribution to this movement and toward the goal of more rigorous and transparent L2 research methods, we asked the following research questions: …”

A Meta‐Analysis of Reliability Coefficients in Second Language Research

“…Using a similar approach with a sample from a different population, Loewen et al. () investigated knowledge of different quantitative concepts and procedures among researchers in applied linguistics. These included, in addition to reliability, terms such as ‘mean,’ ‘null hypothesis,’ ‘standardized score,’ and ‘regression.’ On average, the respondents ( N = 331) rated their ability to interpret reliability as relatively high: M = 4.78 on a 6‐point scale.…”

Methodological Synthesis in Quantitative L2 Research: A Review of Reviews and a Case Study of Exploratory Factor Analysis

“…In particular, the statistical techniques used in SLA “have not developed from within the field” (Loewen & Gass, , p. 181) but are mostly drawn from sister disciplines such as psychology and education. Over the years, SLA researchers have used not only common inferential statistics such as t tests, chi‐square, and analyses of variance but also more advanced statistical techniques, like structural equation modeling and factor analysis (Gass, ; Lazaraton, , ; Loewen et al., ). However, perhaps due in part to its noninnovative nature in terms of statistical methods, SLA lacks field‐specific standards for employing and reporting on such techniques.…”

“…However, perhaps due in part to its noninnovative nature in terms of statistical methods, SLA lacks field‐specific standards for employing and reporting on such techniques. This lack of standards is compounded by the lack of statistical literacy among L2 researchers (Loewen et al., ), both of which merit urgent attention. Although some L2 researchers have called for more rigorous L2 methods, including sound research designs and clear reporting practices (Brown, ; Loewen & Gass, ; Norris & Ortega, ; Plonsky, , ; Plonsky & Gass, ), very little research has investigated how well L2 researchers employ statistical methods to analyze their data.…”