چکیده:
This study intended to determine the way validity and reliability i.e., psychometric properties were reported in the Applied Linguistics research articles. The study also focused on the measurement methods applied to determine the validity and reliability of the scores derived from the tests and questionnaires in the empirical studies. The corpus of the study included 331 empirical studies derived from 733 research articles (RAs) published between 2005 and 2018 in three prominent Applied Linguistics journals – Applied Linguistics, Modern Language Journal, and TESOL Quarterly, The selected papers used test and/or questionnaire for data collection. Our analysis indicated that 77(20.98%) of the studies did not report validity and reliability measures, 82(22.35%) reported only reliability measures, 26(7.08%) reported only validity measures, and 182(49.59%) reported both the validity and reliability measures for the instruments. It was also found that content validity assessed through the pilot study had the highest frequency among validity evidences while internal consistency, mostly identified by Cronbach's alpha, was the most frequent reliability evidence.
خلاصه ماشینی:
Validity and Reliability Reports in Applied Linguistics Research Articles: The case of tests and questionnaire Khalil Tazik, Assistant Professor, School of Medicine, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran khaliltazik@gmail.
It was also found that content validity assessed through the pilot study had the highest frequency among validity evidences while internal consistency, mostly identified by Cronbach's alpha, was the most frequent reliability evidence.
evidence based on consequences which can be assessed based on the unintended use of the instrument and the degree to which it affects the inferred interpretations Chan (2014) also overviewed and presented some approaches and research designs which are used in gathering evidence for validity: "factor analysis, item-test correlations, measurement invariance, differential item functioning, multitrait-multimethod design, item response theory, and experimental and quasi-experimental designs" (p.
Reviewing previous studies on test validation practices, Chinni and Hubley (2014, reported that (1) the frequency of reporting validity and reliability have increased over time (2) the researchers failed to regard characteristics of selected sample during reliability and validity reports according to the previous research (3) Cronbach's alpha by far was more frequent than other reliability estimates (4) validity evidence was limited to some forms and often reported poorly (5) some "validity evidence such as response processes and consequences" (p.
The main reasons for selecting these instruments are (1) their popularity in empirical studies for both large- and small-scale uses (2) their replicability across studies and (3) agreed-upon methods for assessing reliability and validity measures which provides researchers with insights into the underlying constructs of these instruments.