Common sense suggests that interest in a particular topic influences further learningabout that topic. Those with a high interest in mathematics are willing to engage withsuch topic to a larger extent, leading to better achievement. Most researchers in thefield tend to agree with what seems intuitively true: that interest is a cause of learning.However, the empirical data that support this point of view consist mainly of correla-tions between both variables (e.g. Schraw & Lehman, 2001), and correlation doesnot imply causation. In fact, there are at least two other possibilities here, namely thatincreases in knowledge cause interest to increase, or that interest and knowledgeinfluence each other reciprocally.Cross-lagged panel analysis, used in our recent study published in this journal(Rotgans & Schmidt, 2017) and criticised by Hidi and Renninger (Hidi & Renninger,2017), is a methodology that, unlike the use of simple correlations, enables one to testcausalhypotheses about the relationship between constructs. By measuring both con-structstwice,at different points in time, the three directional hypotheses proposed abovecould be tested simultaneously. (The interested reader is referred to the original paperor to the literature on cross-lagged panel analysis for further technical details.) Wefound a directional influence of previously acquired knowledge on later interest, butnotfrom previous interest on later knowledge. These findings, and those of two exper-iments reported in the same article, led us to conclude that, rather than a cause oflearning, interest is a consequence of learning.,
British Educational Research Journal
Erasmus University Rotterdam

Schmidt, H., & Rotgans, J. (2017). Like it or not: Individual interest is not a cause but a consequence of learning. Rejoinder to Hidi and Renninger (2017). British Educational Research Journal. doi:10.1002/berj.3307