Thanks, SteveFoester. Today I received the following assignment: A colleague of yours asked you for your help explaining to a graduate software engineering class why it is inappropriate to discard what appear to be outlying data samples. When asked for the rationale for providing a smaller set of data than would normally be expected given the project protocol, numerous students responded that those data-points were outliers and their elimination improved the correlation they were computing. How would you compellingly explain to these students why this reason alone for exclusion is unethical? The students respond that they have seen the colleague exclude data due to environmental failures (e.g. a power failure during a sorting algorithm performance test) and besides, this is just a class exercise and the colleague should “lighten up?” Is this normal for a research methods class? I DO see the utilitarian use of the methodology, but trying to reconcile my learning of the material with also learning "techno-babble" (for those Star Trek fans, like me) is extremely frustrating. Thinking of transferring.