A common problem in statistics, when researchers deliberately or accidentally ignore results that are not to their liking and only report those that are advantageous. Cherry picked results may mislead people (including the researchers themselves) or make them falsely believe something is true.
Also used in hcistats2e: Chap. 1: page 17; Chap. 6: page 70; Chap. 8: pages 87, 89, 90, 91, 94, 100; Chap. 9: page 107; Chap. 15: pages 187, 188