A Guide to Robust Statistical Methods in Neuroscience

Abstract

There is a vast array of new and improved methods for comparing groups and studying associations that offer the potential for substantially increasing power, providing improved control over the probability of a Type I error, and yielding a deeper and more nuanced understanding of data. These new techniques effectively deal with four insights into when and why conventional methods can be unsatisfactory. But for the non-statistician, the vast array of new and improved techniques for comparing groups and studying associations can seem daunting, simply because there are so many new methods that are now available. This unit briefly reviews when and why conventional methods can have relatively low power and yield misleading results. The main goal is to suggest some general guidelines regarding when, how, and why certain modern techniques might be used. © 2018 by John Wiley & Sons, Inc.

Probabilistic performance estimators for computational chemistry methods: Systematic improvement probability and ranking probability matrix. I. Theory

“…Correlation. The estimation of correlation coefficients by the standard Pearson method is reputed to be very sensitive to the presence of outliers 39 . As the presence of a small amount of outliers is a frequent feature of the benchmarking data sets, we use the more robust rank-correlation (Spearman) method, unless otherwise specified.…”

Prepulse Inhibition of the Startle Reflex as a Predictor of Vulnerability to Develop Locomotor Sensitization to Cocaine

“…For the behavioral sensitization data, a mixed ANOVA was performed with three between variables-Sex (males and females), PPI (High and Low) and Pre-Treatment (Saline and Cocaine)-and one within variable, Minutes (six 10-min periods to analyze locomotor activity in the habituation phase, and four 10-min periods to analyze locomotor activity in the test phase), using Bonferroni's pairwise post hoc comparisons. In order to test our hypothesis about varying vulnerability to developing behavioral sensitization in Low-and High-PPI mice, we also performed the Tukey's honest significant difference (HSD) test, which does not require previous ANOVA with a significant interaction between factors and is highly conservative against type I error (Wilcox and Rousselet, 2018). All results are expressed as mean ± SEM.…”

Intrinsic activity temporal structure reactivity to behavioural state change is correlated with depressive symptoms

“…DFA values at the significant electrodes were then compared between the EC and EO conditions using robust permutation tests of the difference in medians (5000 samples). Outlying values were excluded with reference to the median absolute deviation to the median, with a cut-off of 2.24 (Wilcox and Rousselet, 2018). This analysis aimed to test the reactivity of the intrinsic activity properties to the change in state.…”