Wednesday, July 21, 2010

Nail on head!!!

Namnezia posted today with a quote that sums up everything I have been trying to teach my lab:
Good scientists will gladly let their pet hypotheses evolve or be discarded if evidence shows up to disprove them. Bad scientists will keep looking for specific data supporting their views while ignoring overwhelming amounts of data that don’t support their views.
Grad students, postdocs and everyone who is learning to be a scientist or trying understand what scientists say, take heed!!

7 comments:

ScientistMother said...

sadly, I've seen alot of "bad" scientists. Thankfully, I work for a good one!

Comrade PhysioProf said...

It's even worse than that, because a lab in which the PI is a "bad scientist" in the sense Namnezia describes is extremely fertile ground for scientific misconduct. There should never be a "right answer" that an experiment is supposed to provide.

namnezia said...

I agree with CPP - I think the biases of a PI of what the "right" answer should be will be passed on to his or her trainees, so that even if they are not willfully committing misconduct, their biases will affect their experimental work. For example if a specific assay is not supporting "the hypothesis", well they might say that there must be something wrong with the assay and will discard the data.

It's hard to design good experiments where either outcome, is interesting and leads you to interesting interpretations.

Scientist said...

This is so my PI, we have lost out on publishing 5 papers in the last three years because of this. He is so narrow minded and egotistical, no one's opinion is right although we're the ones doing the work. He will completely ignore us and the science because and I quote "it's my money and I want it done" even if it's bad science. In the end time and money are wasted and actual real data is not believed until it's published by someone else.

Dr.Girlfriend said...

A successful career in science depends on your publication record.

It is unfortunate that certain data will nullify and an exciting hypothesis, or prevent a person completing a story. Scientists will not do certain experiments simply to avoid the risk producing awkward data that will render what they already have as unpublishable or unglamorous. Can we blame them?

In many cases being a good scientist means losing out on a Science or Nature paper because you did not choose your experiments "wisely". It is not misconduct per see, not if you have other good reasons why you did not do certain experiments.

Jade Ed said...

I worked for a PI that told me, when the data I produced was showing that his model was incorrect:
"If you want to prove me wrong, go work somewhere else."

I told him I wasn't trying to prove him wrong. I thought the model just needed to be updated.

I left there on bad terms and about 3 years later, I saw a publication from his lab. The title of the paper? Exactly what I said.

Gerty-Z said...

Couldn't agree more! It drives me bat-shit insane when people talk about the experiment "not working" when it worked perfectly, just didn't give the answer you wanted. But I don't believe that you always have to be "right" to get published. You just need to be able to readjust and fit the new data into a better hypothesis.