Science and replication

[Dave2002]

Here is an interesting article about some aspects of science - publications and replication of published papers - https://www.vox.com/future-perfect/2...iew-statistics

A problem with science and publications is that often people's jobs and promotion prospects are linked to publishing. Quantity not quality may sometimes be the way to get on. The article in question also specifically refers to papers in social sciences, and studies involving experimental sampling methods. Not all published papers are of that form.

Resesarcher bias may tend to push workers towards publications, and as mentioned in the article, results which are marginal at the threshold of statistical limits may still be published, and researchers have an interest in presenting their work in a positive way. In cases like that researchers may qualify their work with comments regarding limitations, though might still tend to do so in a positive way. Pointing out possible limitations is not, in itself, dishonest.

There are difficulties in presenting negative results, though think how many scientific "discoveries" have been made by continuing to investigate after years of negative results. Gravity waves are one immediate example. If scientists hadn't thought that there really was a strong case for showing the existence of such phenomena, they "ought" to have given up their investigations after a few of the initially negative results.

[Richard Barrett]

A problem with science and publications is that often people's jobs and promotion prospects are linked to publishing. Quantity not quality may sometimes be the way to get on.

That applies to humanities just as much as to science.

[french frank]

That applies to humanities just as much as to science.

True. Add as many titles as you can to the end of year research list. And weigh your thesis before submitting. (No, I'm not being sarcastic, though I imagine it refers more to humanities where you're delving into dusty obscurities or avant-garde mystification that no one else knows anything about).

You have to know your own subject thoroughly to be able to judge whether some new paper adds to the tally of human knowledge in that field.

[Frances_iom]

...And weigh your thesis before submitting. .

Maybe engineering is different - I remember sitting in on the PhD viva for the brightest student I ever had - the external walked in very seriously then dropped the thesis on the table with some force - the student was looking very white by then - "right then what half do you want examining on?" at which the student (and I) relaxed and gave, as I expected, an excellent defence of his thesis.

[Dave2002]

Maybe engineering is different - I remember sitting in on the PhD viva for the brightest student I ever had - the external walked in very seriously then dropped the thesis on the table with some force - the student was looking very white by then - "right then what half do you want examining on?" at which the student (and I) relaxed and gave, as I expected, an excellent defence of his thesis.

In some countries, allegedly, the form is to put dollar bills interleaved with the pages of a thesis. Wouldn't happen here, of course.