Regression modelling relationships at chosen percentiles, not just the mean.
Quantile regression estimates how a predictor relates to different points (e.g. 50th, 99th percentile) of an outcome's distribution, rather than only its average. Here it lets the authors show that the bodyweight–strength relationship differs at the top of the distribution versus the median, something a standard mean-based regression would obscure.
This guide was auto-drafted and is pending editorial review.