ABSTRACT
Experimental skills should be acquired by learners at school and university alike. To promote experimental skills, various approaches exist within a spectrum between implicit and explicit instruction. Regarding these instructional approaches, numerous findings are available which predominantly relate to pupils. It is an open question whether it is better to instruct experimental skills explicitly or implicitly in university courses. And if experimental skills are fostered explicitly, what about the content-related skills? Especially in teaching programmes with both experiment-related and content-related objectives, it is questionable whether content-related objectives are achieved to a lesser degree via explicit instruction of experimental skills. Both questions are addressed in this paper. We present an intervention study involving 144 students. All students carry out the same experiments. The intervention groups differ merely in the degree of explicit instruction. Learning gains in experimental and content-related skills are assessed in a pre-/post-test design. The results show that the type of instruction does not significantly influence the acquisition of experimental and content-related skills in this target group. But by trend, the expected differences in learning gains can be observed. This leads to new research perspectives and implications for teaching which are discussed.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Britta Kalthoff http://orcid.org/0000-0002-7117-4994
Heike Theyssen http://orcid.org/0000-0002-6660-1509
Nico Schreiber http://orcid.org/0000-0002-0868-561X
Notes
1 The studies from the meta-analysis carried out by Ross (Citation1988) were considered in the meta-analysis conducted by Schwichow et al. (Citation2016b).
2 Such a question did not, however, form an object of the studies conducted by Koenen (Citation2014) and Schwichow et al. (Citation2016c).
3 The students in the intervention groups also completed a practical experimental task. This means that the test time was extended by 30 minutes. The experimental task was not applied in the baseline group. This aspect will thus not be addressed any further.
4 A total of five items were surveyed for the self-concept (see Brell, Citation2008). Each item was answered on a four-level Likert scale. Cronbach’s α indicating the internal consistency of the scale was .92 across all students.
5 The question of whether marks and self-concept can be assumed to be interval scaled is a matter for critical debate. Equivalent non-parametric procedures indicate lower power. The MANOVA therefore represents a ‘worst case scenario’ for the purpose of monitoring the comparability of the groups.
6 Sensitivity, global effect: N = 144, four groups, six dependent variables, = .05, 1 − β = .80, f2 = 0.05 (calculation using G*Power, Faul, Erdfelder, Lang, & Buchner, Citation2007).
7 If you are interested in the labguides which were used in the present intervention study (in German), please contact the first author.