The Development and Validation of the Self-Efficacy Beliefs about Equitable Science Teaching and Learning Instrument for Prospective Elementary Teachers

Open Access
Ritter, Jennifer Marie
Graduate Program:
Curriculum and Instruction
Doctor of Philosophy
Document Type:
Date of Defense:
June 28, 1999
Committee Members:
  • Thomas M Dana, Committee Chair
  • Peter A Rubba Jr., Committee Member
  • Jeanne Brady, Committee Member
  • Rustum Roy, Committee Member
  • Carla Zembal Saul, Committee Member
  • Science Education
  • Outcome Expectancy
  • Personal Self-Efficacy
  • Riggs
  • Diverse Learners
  • Self-Efficacy
  • Prospective Teachers
  • Bandura
  • Enochs
The purpose of this study was to develop, validate and establish the reliability of an instrument to assess the self-efficacy beliefs of prospective elementary teachers with regards to science teaching and learning for diverse learners. The study used Bandura's theoretical framework, in that the instrument would use the self-efficacy construct to explore the beliefs of prospective elementary science teachers with regards to science teaching and learning to diverse learners: specifically the two dimensions of self-efficacy beliefs defined by Bandura (1977): personal self-efficacy and outcome expectancy. A seven step plan was designed and followed in the process of developing the instrument, which was titled the Self-Efficacy Beliefs about Equitable Science Teaching or SEBEST. Diverse learners as recognized by Science for All Americans (1989) are "those who in the past who have largely been bypassed in science and mathematics education: ethnic and language minorities and girls" (p. xviii). That definition was extended by this researcher to include children from low socioeconomic backgrounds based on the research by Gomez and Tabachnick (1992). First, the researcher defined the self-efficacy construct to be measured and a content matrix for the instrument that identifies the two dimensions of self-efficacy: personal self-efficacy and outcome expectancy and the four diverse learner groups: ethnic, gender, language minorities and socioeconomic status. Second, one hundred ninety-five draft Likert type items, modeled after those composing the STEBI A (Riggs, 1988) and STEBI-B (Enochs & Riggs, 1990) were prepared with at least six representatives for each cell in the content matrix. Third, the draft items were reviewed independently for clarity and comprehension by ten graduate students in science education. Fourth, a panel composed of eight faculty members representing science education, multicultural education, and self-efficacy research independently judged the content validity of the revised items. Fifth, the 48 remaining items were organized as a draft SEBEST instrument. The draft instrument was administered to 226 prospective elementary teachers in the Elementary-Kindergarten Education Program (EK ED) at Penn State in the Fall of 1998. The item analysis was performed with the following goal: What is the most reliable and valid combination of items to compose the SEBEST for the purposes of assessing prospective elementary teachers self- efficacy beliefs for teaching science to diverse learners, and the two dimensions of self-efficacy: personal self-efficacy and outcome expectancy? Sixth, using the results from factor analyses, Coefficient Alpha, and Chi-Square a 34 item instrument was found to achieve the greatest balance across the construct validity, reliability and item balance with the content matrix. The 34 item SEBEST was found to load purely on four factors across the content matrix thus providing evidence construct validity. The Coefficient Alpha reliability for the 34 item SEBEST was .90 and .82 for the PSE sub-scale and .78 for the OE sub-scale. A Chi-Square test (X2 = 2.71, df = 7, p > .05) was used to confirm that the 34 items were balanced across the Personal Self-Efficacy/Outcome Expectancy and Ethnicity/Language Minority/Gender/Socioeconomic Status/ dimensions of the content matrix. In the seventh step the SEBEST was to administered to 102 prospective elementary teachers in the Elementary-Kindergarten Education Program (EK ED) and 23 prospective elementary teachers in the Urban Early and Middle Childhood Education Program (URBED) at Penn State at mid-term, in the Spring of 1999, and administered again to the EK ED prospective elementary teachers at the end of the semester, to gather further information regarding the construct validity and reliability, particularly test-retest reliability. The reliability for the of the SEBEST for the URBED student teachers was .90 for the entire instrument, .81 for the PSE sub-scale and .88 for the OE sub-scale. The reliability for the SEBEST for the EK ED group was .88 for the entire instrument, .83 for the PSE sub-scale and .85 for the OE sub-scale. ANOVA was performed on mid-semester data to determine if the SEBEST could differentiate between prospective teachers prepared to teach diverse populations (i.e., the URBED sample) and those in a contemporary teacher education program that appeared not to have a particular emphasis on teaching diverse populations (i.e., the EK ED sample). The ANOVA results showed no difference between the two groups. While this finding did not provide further evidence of construct validity, it did not distract from the instrument's construct validity. The non-significant ANOVA results were attributed to no difference in self-efficacy beliefs between the two samples, due to the EK ED program having diversity elements throughout the program. Also, one or both samples could have had inflated self-efficacy perceptions regarding science teaching and learning for diverse learners. Test-retest reliability was calculated by applying a Person-Product-Moment to the EK ED data collected at mid-semester and at the end of the semester. SEBEST test-retest reliability was .70. The PSE sub-scale test-retest reliability was .70 and the OE sub-scale test-retest reliability was .67. Based on the standardized development procedures used and the associated evidence, the SEBEST appears to be a content and construct valid instrument, with high internal reliability and moderate test-retest reliability qualities, for use with prospective elementary teachers to assess self-efficacy beliefs for teaching and learning science for diverse learners. Further study of the instruments construct validity is recommended. Norming the SEBEST may provide some insights concerning the test-retest results and will provide additional information on the SEBEST that will be useful to users. Additionally, development of a form of the SEBEST for practicing elementary teachers should be pursued.