Coexistence of Misconceptions and Scientific Conceptions in Chemistry Professors: A Mental Chronometry and fMRI Study

17 chemistry professors (9 female, 8 male; mean age 43; mean 14 years teaching experience)

Pre-university and university chemistry professors in Quebec with at least 5 years of teaching experience - the strongest possible test case, since if anyone has replaced their naive chemistry conceptions it is a person who teaches chemistry for a living.

Very small (n = 17), self-selected, culturally homogeneous (French-Canadian) fMRI plus mental-chronometry study - grade D, and the authors themselves flag that the imaging analysis used uncorrected statistical thresholds and that 31% of stimulus pairs were dropped because the professor answered incorrectly. Its design virtue is the matched-pair construction: 40 pairs of true/false chemistry statements (acid strength, evaporation, molecular properties) identical except for CONGRUENCY - the congruent member involves no plausible interference from a documented misconception, the incongruent member does. Read via full-text (open-access Frontiers).

Chemistry professors were significantly slower on incongruent than congruent statements: 3,631 ms (SD 1,424) versus 3,058 ms (SD 1,196), Wilcoxon z = 7.796, p < 0.001, r = 0.40 (a medium effect). Imaging showed greater activation for incongruent items in left ventrolateral prefrontal cortex, left dorsolateral prefrontal cortex and pre-supplementary motor area - the inhibitory-control network - though not in anterior cingulate as predicted. The authors conclude that misconceptions and scientific conceptions COEXIST in chemistry professors' minds, and that models of conceptual change should abandon replacement in favour of coexistence with context-dependent mobilisation. For a founder the practical translation is blunt: a 573 ms interference cost survives a doctorate plus fourteen years of teaching the correct concept. Whatever conceptual-change instruction achieves, it is not deletion. Note the ceiling this puts on the reading of any post-test: 31% of the professors' responses on paired items were WRONG, in their own subject.

Not applicable - within-subject congruent/incongruent contrast.

Extends Masson et al. 2014 (same group, electricity, expert/novice) to a within-expert design at the extreme of the expertise range, and converges with Shtulman & Valcarcel 2012 behavioural data and with independent fMRI work on physics PhDs.