Statistics for Data Analysis teaches the handful of ideas that stop those mistakes. You will describe a dataset properly, read distributions and outliers, compare groups without fooling yourself, separate correlation from cause, and judge when your sample is simply too small to conclude anything.