You are here: Home / Lectures & Events / Gorski Endowed Lectures / Luke Plonsky (2019)

Luke Plonsky (2019)

Watch it


Speaker: Luke Plonsky

Date of Talk: January 29, 2019

Location: Foster Auditorium

Title and abstract: 

"Methodological reform as an ethical imperative in applied linguistics"

There is a methodological reform movement currently taking root in applied linguistics. Evidence of this movement can be found in a wide variety of outlets such as editorials (e.g., Trofimovich & Ellis, 2015), revised journal author guidelines (Mahboob et al., 2016), methodological syntheses (e.g., Plonsky, 2013), specialized symposia (Norris, Ross, & Schoonen, 2013), studies of methodological literacy and training (Gonulal et al., in press), the newly added Research Methods strand at AAAL, and the introduction of new analytical techniques (Larson-Hall & Herrington, 2010). Several years ago, in fact, Byrnes (2013) referred to what she saw as a “methodological turn” (p. 825) already underway. The concerns that led to and that are often backed empirically within this body of research are not merely fodder for technocrats or nit-picking for research methodologists. To me, arguments in favor of ensuring that we produce “good” research are perhaps best viewed through the lens of ethics.

 

In this paper I will discuss some of the initiatives—ranging from empirical to organizational—currently underway that seek to examine and improve the state of our science. As part of this larger discussion, the paper will also present a study that seeks to complement the body of mostly synthetic evidence of research practices by asking applied linguists about the extent to which they have engaged in a range of questionable data handling and reporting practices including clearly fraudulent activities such as data falsification and fabrication. A survey of questionable research practices (QRPs) based on Fanelli (2009) was sent to 2,230 applied linguists; 364 sets of responses were collected. We calculated descriptive statistics for all item responses. We then compared and correlated responses across and with participant demographics (e.g., years since completion of PhD). The results reveal fairly widespread fraud in applied linguistics: Approximately 1 in 6 participants reported having falsified or fabricated data on at least one occasion. I will discuss this finding, as well as those for other QRPs, from several angles. For example, to provide a broader context, I compare our results to those from other fields and in light of the replicability crisis (see Open Science Collaboration, 2015) currently being discussed in psychology. In addition, I’ll propose a number of possible causes for the prevalence of such practices such as the pressure to publish, insufficient training, and the lack of accountability. I also address the apparently very serious need for methodological reform as a matter of theoretical, practical, and yes—of course—ethical urgency.