It trumpets the gloriousness of Piero Anversa, dumped in disgrace from Harvard for fraudulent research, to which he reportedly admitted; and then, to round out the contemptible behavior, he went ahead and sued Harvard anyway for having damaged his career by, um, having discovered and acted on his lab having – at huge federal government expense, by the way – committed research fraud. (His case against Harvard was dismissed.)
Harvard has to repay the government ten million dollars because of Anversa.
Apparently people in various labs at Harvard knew for over ten years about the guy. Maybe SDSU’s web editors are on the same schedule.
May 1st, 2017 at 9:10AM
Nearly every day for the last year, I have walked by a poster of a research project done by some students in a social science department. They have an attractive graph of their data, with all fifty or so points plotted, and a regression line indicated.
The sad thing is, anyone with some fairly basic math can look at the data points and see immediately that there is no there there. No correlation, no nothing. So what is a student to do? Well, just as your faculty tell you, waterboard the data until it can be made to say something. Chuck enough “outliers,” which aren’t outliers at all, until you can finally get some slope in that regression line. It appears that students no longer are even being inconvenienced with having to come up with a null hypothesis or any of that stuff. After all, your introduction indicates that you are going to “prove” dubious statement X, and you should not let actual results get in the way of that.