Measuring and monitoring conceptions of research

Abstract

This study assessed the validity and reliability of the Meyer, Shanahan, and Laugksch’s Conceptions of Research Inventory using data collected from 227 undergraduate hotel management students in the Netherlands. The results of a series of exploratory and confirmatory factor analyses showed substantial empirical support for the five-factor structure of the scale and for the majority of items used to measure each of the five conceptions. Furthermore, analyses of variances and K-means cluster analyses revealed significant differences among students of three study years in two of the five research conceptions (i.e. Misconceptions and Problem-Solving), suggesting that undergraduates’ research conceptions evolve as they progress through the curriculum. Theoretical and practical implications of the findings are discussed.

Keywords:

• conceptions of research
• cross-validation
• undergraduate hospitality education
• developmental perspective

Acknowledgements

The authors thank Journal Editor Professor Vincent Meek and two anonymous reviewers for their comments and suggestions, Mrs Elizabeth Cutler for her editorial support, Dr Craig Thompson for providing the analytic software, and Professor Jan Meyer for his advice in the writing-up of the paper.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 As a supplementary analysis, we also conducted a series of two-step cluster analyses involving all five research conceptions. We entered all five research conceptions into the analysis. The outcome showed that two clusters were detected – those who scored high vs. those who scored low on all conceptions. This is broadly in line with the inter-factor correlations presented in Table 5 in that all inter-factor correlations were positive and that those who scored high on one factor tended to score high on other factors as well (and similarly, those who scored low on one tended to score low on others, too). In addition, this cluster analysis showed that all clusters fell on year 2 in the cluster comparison analysis. We then began to eliminate factors (or conceptions) to repeat the two-step cluster analysis. We did so by eliminating the factors lowest on the importance scale as shown by the cluster analysis. The outcomes of these analyses still produced two clusters until when only two factors (Problem-Solving and Truth-Seeking) were included. In that case, there were four clusters created. However, all four clusters were concentrated on second-year students. In light of the results of these additional cluster analyses, we are of the opinion that selecting factors on the basis of the previous ANOVA outcome for the K-means cluster analysis instead of choosing factors on the ground of the importance ranking as generated by the two-step algorithm represented the most appropriate approach for these analyses.

2 We calculated the alpha indices of Misconceptions for first-, second-, and third-year students, respectively: first year .68, second year .43, third year .35. These findings suggest that the unidimensionality of this particularly conception decreased as individuals’ research knowledge grew more sophisticated. The implication here is that the items do measure research misconceptions reliably. It is just that misconceptions grew into different directions as students’ research-related knowledge grew more sophisticated.