A meta-analytical study of the effects of virtual and hands-on laboratory work on affective learning outcomes in the natural sciences

Innovations in the field of natural sciences have significantly improved the quality of modern everyday life. Yet, students commonly have a negative attitude towards the natural sciences due to the abstract nature of the subject content and the receptive approach to learning. Affective learning outcomes in this field can be improved through both hands-on and virtual laboratory work. However, previous (quasi-)experimental studies comparing their effectiveness in this regard have yielded highly conflicting results. To better understand which laboratory work type more positively influences affective learning outcomes in the natural sciences, this meta-analysis synthesized the results of 22 pertinent (quasi-)experimental studies published in the past 20 years. Virtual laboratory work proved to have statistically significantly higher effectiveness. Furthermore, the meta-regression showed that virtual laboratories’ effectiveness was moderated by the approach to experimental work and the type of Likert scale used to evaluate affective outcomes. The results revealed a significantly higher effectiveness of guided versus open inquiry and a significantly higher mean g value when using a four-point versus a five-point Likert scale. These findings point to fruitful approaches to improving affective learning outcomes in the natural sciences, along with modes of evaluating their achievement. They also highlight the necessity of devoting special attention to the development of instruments for evaluating the effects of different laboratory activity types on affective learning outcomes. Further research is needed to examine the role of open versus guided inquiry in achieving these outcomes.