UBC Theses and Dissertations
Development of an ethnocentrism scale for junior high school students in British Columbia Thaler, Carol-Lyn Sakata
Ethnocentrism is a cognitive bias whereby people and practices of other cultures are viewed and judged according to what is right and wrong in one's own cultural group. Social studies teachers, when teaching about peoples and cultures, implicitly, if not explicitly, try to prevent the development of an ethnocentric attitude in their students. The purpose of this study was to develop a valid and reliable ethnocentrism scale for use by social studies teachers to gain feedback of how a particular unit of study effected their students. The scale can be used as a pre-post test measure before and after a course or unit of study. The known California E-scale, British Ethnocentrism Scale, and Australian Ethnocentrism Scale became the models for the construction of this scale. The developed ethno-centrism scale is a 30 item Likert summated rating scale using six response steps ranging from "agree very much" to "disagree very much". To insure content validity, test items were based on interviews with members of several minority groups in the Lower Mainland of British Columbia to gather information about practices and customs particular to their culture. The whole study was conducted using grade eight and nine students from four junior high schools in one Lower Mainland School District. The pool of test items were administered to approximately 550 subjects to gain data for item and factor analysis. Two ethnocentrism scales were constructed using items with high reliability rankings. The two forms were administered to groups of students to establish whether the forms were statistically parallel. The forms were not statistically parallel. One form using items with high item reliability rankings was constructed. An intervention study was conducted to check for construct validity. Eight social studies classes in one school were divided into experimental and control groups. The experimental classes were exposed to a three day lesson on ethnocentrism. The developed scale was used as a posttest measure. A significant difference between the two groups was reported. The experimental group means were significantly lower (p < .05), less ethnocentric, than control group means. Finally, the scale was administered to 215 students to gather data for test-retest reliability. At this stage the students were also administered a dogmatism scale and a self concept scale to check for concurrent and construct validity. Test-retest coefficent was high (p = .83) and the Hoyt reliability coefficent for test consistency was high for all administrations of the form ranging from .87 to .94. Recommendations for use of the developed ethnocentrism scale and areas for future research were based on the findings.
Item Citations and Data