OT Rant: I love neuroscience but hate statistics

[likemike]

I'm a soph undergrad. I started out w/ psychology, but I just didn't like the classes. I became very interested in neuroscience, however, I really dislike statistics. I find it difficult but also have a "what's the point?" attitude sometimes. Surprisingly, I am not turned off by neurobiology (since I never really cared for biology in high school).

I thought I would get involved with experimental psych and cognitive neuro, b/c I was interested in emotion, decision making, memory... but I feel more comfortable with the methods, techniques, and metrics of neurobiology (cell cultures, proliferation, cell migration).

Admittedly, I took psych stats my first semester and didn't do very well (since I was goofing off a lot). But I feel like it's not just lack of practice, but perhaps I'm fundamentally suited for something else. But perhaps I'm naive to think that stats are not also going to be very critical in neurobiology.

I'm currently contemplating which direction I should go for research assistant positions this summer...

---

Members don't see this ad.

 

[Neuronix]

Stats for Psychology students is pretty basic and not that hard. Overall, I think you're kind of complaining about nothing. You should suck it up if you really like neurobio and put up with those few requirements that aren't exactly your pot of tea. That's life. Besides, stats are an important skill for understanding clinical research relevant to practice anyway.

As for research assistant positions, pick one with a good-hearted and friendly mentor who understands what you're trying to accomplish. All other things are secondary. Of course you should be interested in the work, but in reality, you're an undergrad, if you find a good mentor, convince yourself the work is interesting and go at it full steam.

Oh, you probably want to set your user status to pre-med. Psychology student means a clinical psychology PsyD/PhD student. You're going to make everyone think you're pre-that

 

[itsallthesame]

a small something to add... stats was the lamest class when i took it first semester without really understanding the application to research. once you actually get your feet wet in a lab, running experiments, reading a lot of papers with shady data, etc you'll start to appreciate a fine understanding of statistical methods. I'd highly recommend a quick survey of stats before going into the lab but then diving in deeper once you get a feel for designing experiments yourself. It's incredibly useful to know how to make appropriate comparisons of data both to analyze your own results and to see how often people try to slide marginal data through by using tests that may not have as much power as they claim. It may seem lame, but you'll wish you'd paid more attention when you have to write your first paper, thesis, whatever.

 

[Vader]

There are numerous published papers in highly-regarded journals that use questionable or incorrect statistics, including misuse of tests, failure to correct for multiple comparisons, using too small of an "n", wide confidence intervals, etc, that limit the interpretation of the presented results. Without a good understanding of these facets, a reader might incorrectly assume that the statistics are good by the fact that the article has been peer-reviewed and published.

Also, statistics become extremely important for designing your own studies, analyzing the data, and interpreting the results. Most papers in neurobiology nowadays use some element of statistics (i.e. comparison of knockout to wildtype, treatment to control, etc). Qualitative data is fine and often quite impactful, but generally needs to be supported by hard, quantitative data analyzed by rigorous statistical methods.

I remember that I also found general statistics somewhat dry, but when I learned biostatistics and how this applies to research, the subject seemed much more interesting and important to me.

 

[anemone2]

I thought I would get involved with experimental psych and cognitive neuro, b/c I was interested in emotion, decision making, memory... but I feel more comfortable with the methods, techniques, and metrics of neurobiology

I think I know where you're coming from here. I'm also interested in the questions of experimental psych/cog neuro, but can't see myself enjoying the day-to-day stuff. It seems like cog neuro (thinking imaging or modeling here) makes much more extensive use of statistics, computer programming, and other things computational that put some people off. Also, it's a lot less hands-off than neurobio - seems like you spend most of your time at a computer (as opposed to a bench, haha).

But perhaps I'm naive to think that stats are not also going to be very critical in neurobiology.

When it comes to statisics, I think overall neurobio is a bit less demanding - no crazy mathematical models here - but you'll still definitely need to know the basics taught in a freshman stats course..

I'm currently contemplating which direction I should go for research assistant positions this summer...

If you're planning to pursue a research career, I would give both directions a chance (if you have the time). I've always regretted never having given cog neuro a shot, just to be sure I don't enjoy it as much as I think I won't (if that makes any sense..)