Iranian TEFL Graduates’ Conceptions of Measurement Error in Research: A Genealogical Narrative Inquiry

Abstract

The aim of this study is to investigate Iranian TEFL graduates' conception of measurement error in research. Adopting a sequential explanatory multi-method strategy (Borg, 2009), the researchers analyzed causal and temporal relations in the research narratives elicited from 30 TEFL graduates. Gee's (1986) framework for identifying narrative discourse units (lines, stanzas, and episodes) was adopted to investigate participants' conceptions of logical orders in measure development algorithms and their knowledge of error sources. In addition, taking a narrative positivistic approach, the narratives were rated based on Optimal Matching Analysis (OMA). Finally in ‘continuous event history modeling' phase of the study, Cox Proportional Regression Analysis showed how temporal markers in research narratives can be used to predict one's knowledge of measure development in research design. The results suggest that researchers' error-awareness and algorithmic knowledge correlate significantly with each other and constitute knowledge of measure development in general. The contribution of dimensionality and validity testing to this knowledge was also found to be statistically significant.