Item Development and Psychometric Evaluation of the Scrambled Sentences Task: A Pilot Study in Persian

Asiyeh A'lam Hakkakan; Setareh Mokhtari

doi:10.18502/ijps.v20i3.19043

Asiyeh A'lam Hakkakan Institute for Cognitive and Brain Sciences (ICBS), Shahid Beheshti University, Velenjak, Tehran, Iran.
Setareh Mokhtari Institute for Cognitive and Brain Sciences (ICBS), Shahid Beheshti University, Velenjak, Tehran, Iran.

DOI: https://doi.org/10.18502/ijps.v20i3.19043

Keywords: Adaptation; Depression; Interpretation Bias; Linguistics; Psychometrics; Scrambled Sentences Task

Abstract

Objective: One significant concern regarding the cross-cultural use of psychological instruments is their adaptation to the language of the target population. The Scrambled Sentences Task (SST) exemplifies this issue. As a well-known paradigm for assessing interpretation bias (IB), the SST has been widely used across different languages; however, it remains unclear whether the SST is a valid and reliable tool to be used in languages other than English. The current study aims to develop SST items for Persian language while controlling for word frequency, word length and morphological complexity— linguistic features critical to meet SST’s unique demands. We also seek to examine the psychometric properties of the Persian version of the SST (P-SST).

Method: The verbal stimuli for the P-SST were developed by drawing on a pool of sentences from prior research as a guide. These sentences were adapted specifically to fulfill the task's unique demands, ensuring the appropriateness of the P-SST for detecting IB. Since the SST primarily measures depressogenic tendencies, the Beck Depression Inventory-II (BDI-II) was also employed as part of the assessment. The measures were administered to a sample of 222 female students, selected due to evidence suggesting higher susceptibility to depression among women. The reliability and validity of the P-SST were then calculated, along with an analysis of responses to identify areas for enhancing performance on this task.

Results: Our results provided evidence of the convergent validity of the P-SST, as it was significantly correlated with the BDI-II (rs = 0.43, 95% CI [0.29–0.57], N = 161), as well as evidence of its divergent validity (rs = 0.35, 95% CI [0.22–0.49], N = 161). Moreover, internal consistency analysis revealed a Cronbach’s alpha value of α = 0.81 and a split-half correlation value of r = 0.9.

Conclusion: Findings from this research established the psychometric properties of the P-SST as a quick and easily administered assessment tool to be used in the context of depression in Persian. The paper concludes with both linguistic and methodological recommendations to guide future development of SST items in any language.