Monday, May 17, 2010

Sometimes there just is no difference

In our Masters program, the exams are based on figuring out the experiments that have been published, and making conclusions from them.

It's very important to include figures where there's no difference between mutant and normal cells. Because, I swear, 3 of 4 students will insist there is a difference, and one that fits with the expectations generated by other experiments. Give them a figure with a difference and they'll usually identify the right one. Usually you can't miss it. But getting them to conclude that two results are the same, or that it isn't clear, is another question.

It's okay to say there's no conclusion to be made. In this particular figure you just can't tell.

Since most of the experiments we do give results that are not different between the test and control groups, it's essential to be able to say No; as much as we'd really like to say Yes.



Argent said...

We always want a answer though, don't we, as if 'no conclusion can be drawn' is not one.

Reya Mellicker said...

Love the graphic. It looks like a snow fence in western Colorado.

OK. Weird tangent, I'll agree.

Love your new banner, too.

EastwoodDC said...

Also, there might be a difference, but it is so small that we don't care about it. Statistical significance does not imply scientific/clinical/practical significance.

And it does look a bit like a snow fence good thought Reya!

NanU said...

Indeed, though that's a whole new topic. At this point I just want them to be able to say there's nothing there, instead of inventing something.