Dispelling the Myth: Training in Education or Neuroscience Decreases but Does Not Eliminate Beliefs in Neuromyths
Macdonald K, Germine L, Anderson A, Christodoulou J, McGrath LM · 2017
grade Dreviewindependentreplicatednumbers spot-checked
Sample
3,877 respondents: 598 educators, 234 high-neuroscience-exposure, 3,045 general public
Population
United States adults recruited online via TestMyBrain.org, Aug 2014-Apr 2015; educator subsample spans general education (31%), early childhood (28%), higher education (27%), special education (22%)
Design
Large cross-sectional survey benchmarking educator belief against the general public and against people with substantial neuroscience exposure. Recorded as 'review' for the same vocabulary reason as Dekker 2012. Self-selected online sample; groups differed on age, gender and education, which were carried as covariates. Instrument adapted from Dekker 2012 - answer format changed to true/false, two items dropped (caffeine, fatty acids), two added (Mozart effect, dyslexia). Only complete responses analysed.
Key findings
Neuromyth endorsement was 68% in the general public, 56% among educators, 46% among the high-neuroscience-exposure group (all pairwise differences p<0.001 after covarying age, gender, education). Seven 'classic' neuromyths loaded on one factor - two learning-styles items, dyslexia-as-letter-reversal, the Mozart effect, sugar and attention, right-brain/left-brain learners and the 10%-of-the-brain claim. BOTH of the two most endorsed myths in every group were learning-styles items. Training reduces but does not eliminate the beliefs - and does not touch the learning-styles item at all, where the high-neuroscience-exposure group (78%) did not differ from educators (76%).
Genetic confound
Not applicable; outcome is respondent belief.
Replication notes
Itself a conceptual replication of Dekker 2012 on a US sample using an adapted version of the same 32-item instrument. On the three items worded identically, US educators scored modestly better (learning styles 76% vs Dekker-UK 93% / NL 96%; sugar and attention 50% vs 57%/55%; 10% of brain 33% vs 48%/46%), which the authors attribute to recruitment differences (online self-selection vs school-based invitation). The pattern - high prevalence, learning styles top - is what replicates.
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Neuromyth endorsement | mean percentage endorsed | general public 68%, educators 56%, high neuroscience exposure 46% (7-item factor; all pairwise p<0.001) | self-report survey | cross-sectional | none | not-applicable | non-cognitive |
| Learning-styles meshing myth (item 14): "Individuals learn better when they receive information in their preferred learning style" | percentage endorsing | general public 93%, educators 76%, high neuroscience exposure 78% - most endorsed item in every group | self-report survey | cross-sectional | none | not-applicable | non-cognitive |
| Sense-dominated learning styles (item 26): "Children have learning styles that are dominated by particular senses" | percentage endorsing | general public 88%, educators 71%, high neuroscience exposure 68% - second most endorsed item in every group | self-report survey | cross-sectional | none | not-applicable | non-cognitive |
| Neuroscience training as protection, item-level (result cutting against the paper's own title): learning-styles items | educator vs high-neuroscience-exposure contrast | no significant difference on either learning-styles item (76% vs 78%; 71% vs 68%; shared superscript b). Training separates the groups on 10%-of-brain (33% vs 14%) and left/right brain (49% vs 32%) but not on the myth that matters most for classroom practice | self-report survey | cross-sectional | none | not-applicable | non-cognitive |
| Size of the neuroscience-training benefit among educators | partial eta squared / raw item count | eta_p2 = 0.035; educators with several neuroscience courses endorsed 1.3 fewer myths out of 7 than those with none - the authors themselves call this 'modest' (figures as reported by Rousseau 2021 from this dataset) | self-report survey | cross-sectional | none | not-applicable | non-cognitive |
Cited by
- Does matching instruction to a child's "learning style" improve learning?debunkedconf: highgc: low