[conjectures]

Really interested in this topic, not sure about this paper:

- Downplays how inaccurate the classifiers are (55-60% accuracy according to the figures).

- No table of actual left / right frequency pushing frequency. So we can't compare this to the empirical rate of chance (they assume 50% - I would bet £100 that it's not 50%).

- Exclusion policy of people who don't push buttons the right way was inadequately justified, particularly given the above.

- Vague explanation of their statistical methods, even in the supplementary material. In particular I can imagine several ways to interpret what they said about the ANOVA - it shouldn't be up to me to guess what they did, the paper should tell me.

- No use of hold-out data sets.