Data for assessing EA and knowledge in selected teacher colleges in Tanzania

Published: 4 January 2023| Version 2 | DOI: 10.17632/52342x45rv.2
Josephat Paul


This dataset was collected in response to the exploration of the intervention conducted to assess the effect of inquiry-based learning on environmental attitudes and knowledge of pre-service biology teachers in selected teacher colleges in Tanzania. This dataset contains pre and posttests for two colleges and posttests only for other two colleges adding up to four teacher colleges from which data were collected. This dataset provides an overview that lays the foundation upon which further research can be carried out to enhance environmental education and sustainability education in teacher colleges in Tanzania and beyond other areas that bear similar contexts.


Steps to reproduce

These data were collected using three key instruments. First, the two-major environmental values (2-MEV) was employed to measure environmental attitudes based on a two-factor model of Preservation (biocentric view) and Utilization (anthropocentric perspective). Second, a competence model of environmental education was used to measure environmental knowledge on three dimensions of system knowledge, action-related knowledge, and effectiveness knowledge. Third, the data collected were controlled for socially desirable responding using the Balanced Inventory of Desirable Responding short form (BIDR-16). To assess the impact of the intervention, a modified Solomon four-group design was used. Specifically, two groups one from the experimental group and the other from the comparison groups received pre and post measurements whereas the remaining two groups received post-test only. The motive of this design was to control for pre-test sensitization and maximize internal validity of the collected data and of the resulting outcomes. Analysis was conducted to assess the impact. Each group was analysed separately to avoid pseudo-replication by comparing mean scores of the pre and post test using paired samples t-test for normal distribution and Wilcoxon signed-rank test for non-normal data. Moderation effect of socially responding was assessed to examine if the relationship between environmental knowledge and environmental attitudes depended on the amount of socially desirable responding. This was done using the Hayes’ PROCESS macro (version 4.1) in the regression analysis. Therefore, the resulting dataset that is presented here originates from the whole of this process.


University of Dodoma College of Education




the World Bank through the African Centre of Excellence for Innovative Teaching and Learning Mathematics and Science (ACEITLMS)

ACE II-P151847