Clement on embodied examples that don't work

For my other project, I'm reading Clement's paper on "anchoring conceptions" and "bridging analogies":

J. Clement, D.E. Brown, and A. Zietsman, “Not all preconceptions are misconceptions: finding ‘anchoring conceptions’ for grounding instruction on students’ intuitions,” International Journal of Science Education 11, no. 5 (1989): 554–565.

In this paper, they look for "anchoring examples," which are physical scenarios for which at least 70% of students give a correct explanation with high confidence before instruction. The idea is that these scenarios have the potential to be used in instruction to help students transfer their correct intuition to more difficult examples.

I was particularly struck by his discussion of physical scenarios that he expected to be intuitive students because they can easily put themselves into the scenario, but that turned out not to be so intuitive:

There were also some cases for which we mistakenly expected certain anchors to be stronger than others. For example, given the situation of a hand pushing down on a spring in question 4, students were asked whether the spring exerts a force on the hand. This was considered to be a good candidate for an anchor, but we had some reservations about how strong an anchoring example it would be. We expected that the upward force would be recognized more intuitively in the case of holding up a 30 pound dictionary on an outstretched hand (question 2). In both cases the subject can imagine feeling the upward force, but the dictionary situation involves a person exerting the force and allows for direct use of kinesthetic intuition. However, the results indicated that the hand-on-spring situation was in fact an anchoring example for more students (belief score of 80%) than the dictionary-on-hand situation (belief score of 65%). One possible reason for the spring being a stronger anchor is that the spring moves up when the hand is removed, whereas this is not so obvious for the hand when the book is removed.

Perhaps the most surprising result from this study was the low belief score for the log exerting a force on Mr T’s chest in question 1. We predicted this situation would be a solid anchor for students because of the opportunity to identify with the person in the problem. However, this situation was an anchor for only 53% of the students. A full 30% answered, although some with low confidence, that the log would not exert a force. We are interested in using deeper probes and analysis techniques to determine the origins of these anomalous responses in the future.

His description of what he expected seems very much in line with the way we describe how embodied learning activities build on the theory of Ochs. He does not offer any speculation about why these anchors didn't work as well as he expected. Surely this has some kind of implications for our work...