Developing and validating EFL items

This paper presents and discusses the results of a pilot study carried out to validate two English as a Foreign Language (EFL) tests intended to measure implicit and explicit knowledge of the simple past tense (form and meaning). The tests were developed as part of research project which investigates the extent to which immediate electronic corrective feedback (e.g., Heift, 2001, 2004, 2010; Moreno, 2008; Nagata, 1993; Nagata & Swisher 1995; Pujolà, 2001; Rosa & Leow, 2004; Sachs, 2010; Sanz, C., Lin, H. J., Lado, B., Wood Bowden, H., & Stafford, C., 2009; Sanz & Morgan-Short, 2004) given to EFL students carrying out Information and Communication Technology (ICT) based learning activities can affect foreign language development. In particular, the project seeks to understand whether certain types of corrective feedback are more effective than others and whether the effectiveness of corrective feedback might vary depending on the type of task. As this main study will employ a pretest-posttest design with treatment, it was necessary to create and pilot the two tests before they could be used as data collection tools to measure the effects of the treatment provided. The piloting of the tests was therefore carried out in order to validate whether the tests are comparable and understand the extent to which they are suitable for their purpose. This paper begins with a brief outline of the main research project. The core of the paper then reports on the pilot study carried out, focusing in particular on an analysis and critical evaluation of the content of the tests, and provides suggestions on how the tests can be improved in order to guarantee validity and reliability.