The Effectiveness of the Extended Forced-Choice Format as an Alternative to the Likert Scale Format in Improving the Factor Structure of the Psychological Scales

Abstract

The study aimed to examine the relative effectiveness of using the extended forced-choice format as an alternative to the Likert scale format in improving the factor structure of the psychological scales. The neuroticism factor, which is one of the five factors of the Jordanian version of NEO personality Inventory was used as an example of the psychological scales. The Descriptive correlational approach was used to achieve the objectives of the study. The data were collected from (532) available undergraduate male and female students of different academic levels and different majors who are enrolled in different sections and different colleges in four Jordanian Universities, where the selected sections were randomly assigned into two groups (Likert format, and extended forced-choice format). The results of the exploratory as well as the confirmatory factor analyses of the scale in each of its two forms revealed that the extended forced-choice format produced better construct validity indices compared to the Likert scale format, where it was found that the factor structure of the Likert scale format is contaminated; that is, the scale does not measure a unidimensional trait as it is supposed. The results also indicated a better stability index (Cronbach-Alpha) for the extended forced-choice format which was found to be (0.847) compared to the stability index of the Likert scale format which was found to be (0.705). It was recommended to conduct more studies on using the extended forced-choice format as an alternative to the Likert scale format in other psychological scales, especially if the scale includes reversed worded items besides positively worded items.