60 Factors associated with differences in digitally measured student perceptions of teaching quality 3.5.4 Limitations of the study and suggestions for future research In this research, 26 mathematics teachers used the Impact! tool for 4 months in their Grade 9 classes to obtain feedback from students about the quality of their teaching. We consider it important to conduct further large-scale research, to have more statistical power, and to be able to investigate to what extent our findings can be generalized to similar contexts (Cohen, 1988). We also do not know whether our findings can be generalized to other educational contexts, for example, to other subjects than mathematics, and to other grades and other educational levels (such as primary schools or higher education). Such educational contexts ideally will be included in future research. Furthermore, we acknowledge the limitations associated with the instrument used to measure teaching quality. The high correlations between the characteristics of effective teaching might well be due to issues of social desirability (especially for correlations higher than 0.90). We are also aware of the small number of items used to measure most of the teaching characteristics and the over-representation of the clear-instruction items. The instrument will be improved by adding more items to the teaching quality scale and by redesigning the clear-instruction items. We also recommend the investigation of the relationship between students’ teaching quality ratings and more objective teaching quality ratings (e.g., by means of classroom observations by experts). By doing so, causal relationships between the variables investigated can be examined. The relationship between students’ teaching quality ratings and their learning outcomes could also be examined in future studies (the predictive validity of the student ratings): Do students of teachers with high teaching quality ratings learnmore than students of teachers with lower ratings? In future research, potential interaction effects (e.g., between teacher age and teaching experience, and between teacher likeability and classes’ average mathematics grade) could be studied, and researchers should try to ensure greater ethnic diversity and variability in the size of the classes in their samples. The relationship between students’ teaching quality ratings and other independent variables, such as teachers’ personality traits (e.g., teachers being more child-friendly or more no-nonsense), or their educational levels (e.g., academically trained teachers versus teachers with other prior working experience than teaching), or students’ socioeconomic backgrounds could also be investigated in future research. These factors were not included in the current study. By including them, differences in digitally measured student perceptions of teaching quality could be explained even more completely.
RkJQdWJsaXNoZXIy MjY0ODMw