Urdu Translation and Adaptation of Cornell Critical Thinking Test-Level Z in Pakistan

Abstract

This study aims to translate and adapt the Cornell Critical Thinking Test- Level Z (CCTT-Z) into Urdu language in Pakistani population. Phase-I of the study dealt with Urdu translation, back translation, and cross-language validation. Recommended protocols were followed for forward and backward translations. Results of cross-language validation based on a sample of college students (n=42, Mage=18.69 years) showed that scores on both Urdu and English language versions were positively correlated (r =.31, p<.05). Moreover, 45.2% of students reported the Urdu language version easier than the English version. Certain lingual changes were made to make the test more culturally compatible without disturbing its core structure. For the Phase-II, a sample of 367 students of undergraduate and graduate programs (Mage= 21.56 years) was obtained from colleges, universities, and a teaching hospital. The results showed that the test had a high difficulty (p=.38). Twelve items had non-significant point-biserial correlation coefficients and dropping them improved overall Cronbach alpha reliability of the test (from ?=.47 to ?=.59). Five-factor theoretical model and four-factor model (induction items removed) were examined through CFA. The test showed poor model fit and did not appear to have any meaningful factor structure. This finding was not unique as several other validation studies across various cultures also do not show theoretically predicted factor structure. The findings are discussed by examining the available literature on the cross-cultural validity studies done in several other countries.