How to Enhance the Power to Detect Brain–Behavior Correlations With Limited Resources

Abstract

Neuroscience has been diagnosed with a pervasive lack of statistical power and, in turn, reliability. One remedy proposed is a massive increase of typical sample sizes. Parts of the neuroimaging community have embraced this recommendation and actively push for a reallocation of resources toward fewer but larger studies. This is especially true for neuroimaging studies focusing on individual differences to test brain–behavior correlations. Here, I argue for a more efficient solution. Ad hoc simulations show that statistical power crucially depends on the choice of behavioral and neural measures, as well as on sampling strategy. Specifically, behavioral prescreening and the selection of extreme groups can ascertain a high degree of robust in-sample variance. Due to the low cost of behavioral testing compared to neuroimaging, this is a more efficient way of increasing power. For example, prescreening can achieve the power boost afforded by an increase of sample sizes from n = 30 to n = 100 at ∼5% of the cost. This perspective article briefly presents simulations yielding these results, discusses the strengths and limitations of prescreening and addresses some potential counter-arguments. Researchers can use the accompanying online code to simulate the expected power boost of prescreening for their own studies.

No effect of age on emotion recognition after accounting for cognitive factors and depression

“…Although it is possible that this finding may reflect the true state of affairs, it is worthy of further investigation due to the fact that the current sample had a limited distribution of high alexithymia scores, with only 2.9% of the sample scoring above the threshold indicating the presence of alexithymia, and even those scoring above cutoff did not have especially extreme scores. Therefore, it is possible that the limited distribution of alexithymia scores may have contributed towards reduce power in detecting a unique effect of alexithymia on the relationship between age and emotion recognition, as well as a relationship between alexithymia and interoceptive accuracy 1 (e.g., de Haas, 2018). As well as a limited distribution of scores, it is also possible that the use of the heartbeat counting task as a measure of interoceptive accuracy may have contributed towards the lack of a unique effect of interoceptive accuracy on emotion recognition.…”

Two distinct feedback codes in V1 for ‘real’ and ‘imaginary’ internal experiences

“…This suggests a link between mental imagery ability and the precision of imagery-related visual cortex signals measured by fMRI. To increase the probability of finding meaningful imagery–related V1 activity patterns and thereby increase the statistical power of the study( 28 ), we ran behavioural pre-screenings to identify individuals with good visual imagery abilities. To do so, we used a behavioural paradigm that quantifies individual imagery by measuring its impact on subsequent conscious perception of a binocular rivalry display( 11, 24–26 ).…”

Individual differences in science competence among students are associated with ventrolateral prefrontal cortex activation

“…The sample size is lower than the recommended group size ( N = 20–24) to obtain 80% of statistical power in fMRI studies (Desmond & Glover, ) owing to ethical regulations, due primarily to the difficulty in recruiting minor age participants and obtaining school boards authorizations and parental consent. However, findings from simulations suggest that behavioral prescreening and the selection of extreme groups can be a more efficient way of increasing power than increasing sample size (de Haas, ). For example, “prescreening can achieve the power boost afforded by an increase of sample sizes from n = 30 to n = 100 at 5% of the cost.” (de Haas, , p. 1).…”

“…However, findings from simulations suggest that behavioral prescreening and the selection of extreme groups can be a more efficient way of increasing power than increasing sample size (de Haas, ). For example, “prescreening can achieve the power boost afforded by an increase of sample sizes from n = 30 to n = 100 at 5% of the cost.” (de Haas, , p. 1).…”