Evaluation of Five Guidelines for Option Development in Multiple-choice Item-writing Evaluation of Five Guidelines for Option Development in Multiple-choice Item-writing Evaluation of Five Guidelines for Option Development in Multiple-choice Item-writing

Non-profit academic project, developed under the open access initiative Multiple-choice items are used extensively in classroom achievement tests. This explains the importance attributed to understanding how to write these items correctly. Multiple-choice test items may be presented in a question or completion format, with negative, positive, simple or complex sentences, with language that is more or less suited to the subjects being tested, with a set number of options, presented vertically or horizontally, and with different lengths or grammatical characteristics. The adequacy of these aspects was considered by different guidelines importance of these aspects for the quality of test items is an assumption based more on common sense than on the findings of well-founded research. Apart from studies on the number of options, the amount of empirical studies concerning item-writing guidelines is clearly insufficient as has been stated by Haladyna and Downing (1989b), Haladyna et al. (2002). The present study deals with five of the guidelines on options development included in the above-mentioned studies. These guidelines are: i) Keep all options homogeneous in content; ii) Avoid specific determiners; iii) Keep the length of options about equal; iv) Avoid «All of the above» option; and v) Avoid «None of the above» option. This study examines a) test-maker non-compliance with these guidelines and b) the influence this may have on item quality, using a broad set of items taken from university classroom achievement tests that were written without these guidelines being considered. The study's first objective was to evaluate the frequency of errors made by test-makers who did not consider these guidelines, and their distribution between correct and incorrect options., where the sample of items allows us to do so, this study evaluates whether the statistical indexes of difficulty and discrimination vary depending on compliance or non-compliance with the stated guidelines. In addition, it adds two other aspects. On the one hand, it studies whether these item indexes vary depending on whether non-compliance occurs in a correct or incorrect option. On the other, it analyzes whether non-compliance with one of these guidelines in the correct option affects equally the proportion of subjects that choose it in the high and low competence groups. Method Participants The sample of participants in this study came from a population of first-year undergraduate students in Psychology at the This paper evaluates certain guidelines for writing multiple-choice test items. The analysis of the responses of 5013 subjects to …