in that case, it's a question of proportion. we cannot automatically conclude that a (supposed) "good handful" doing good research makes up for "most students" doing bad research.