As a peer reviewer of a number of published academic papers in transportation, my main concern is the poor state of the peer review process, where any really serious criticism of really poor science, data, or attempts at data manipulation in many really bad papers, more often results in replacing the peer reviewers, than in pulling the often misleading and really bad papers. These papers then still seem to often get published, anyhow, and then are often cited...... as somehow representing a "finding" of fact.

I found several papers during my career where the data and analysic clearly did not support the "finding" or "outcome" purported in the paper. I had a grad student in near tears once when I pointed this problem out.... This poor kid was trying to win "approval" from a graduate advisor who had stated a contradictory view, to that supported by the actual data this student had collected, and he was terrified, even though the advisor was clearly wrong.

In another case (out of UC Berkeley, no less) I found that when a certain oft cited researcher could not get the conclusion they claimed from the data, that researcher simply excluded a major part of the part of the data that caused the problem (because it contradicted the researchers preferred conclusion), for no apparent valid statistical reason that could be found. And then, as the old cartoon goes, "the miracle occurred....".

True believers, house whores responding to funders dollars, and careerist's or young grad students who desire to win the "approval" (for tenure, political correctness, outside funders and contracts) of their department, or the tyranny of a grad advisor appear to be among the most egregious causes of this kind of dubiously "scientific" behavior that I have seen, for this kind of nonsense.

Caveat Emptor......