/sci/ - Science & Math

>>4839410
Good choices! Way better than the degree I graduated with.

Where are you going after undergrad? Are you doing any research?

>>4839379
Stony brook represent.

>>4839377
Clinical psychiatrists ARE NOT striving to understand the human condition. They are M.D.s interested in altering behavior with medication and improving the quality of life of people with psychopathology. People interested in understanding the human condition are researchers in the social and bio/medical sciences.

Not a single thing in the world could convince me that a psychology undergraduate degree is required for psychology research of any kind or practicing clinical psychology.

I majored in psychology in college because I thought it was interesting. After taking a few classes, I concentrated my efforts on math and programming and ran like fucking hell into a biology graduate program.

After taking the classes, I can assure you that the most important messages for a clinicial are summed up concisely in an MSW program, and that the details of executing experiments in the social sciences are the same as experimentation in any other field (save for preference for a few obscure statistical tests and a tolerance for higher p-values). It *is not* a useful or necessary degree, even if you plan on pursuing a career in Psychology.

Go ahead, refute me. I dare you.

Neuroscience Ph.D student, first year. dunno yet, fall is probably something like

Principles of Neuroscience
Neurobiology
some electrophysiology lab rotation

starting in a monkey lab on Monday, though. :3

>>4667855
Yes. Psychotropic drugs fit into these: http://en.wikipedia.org/wiki/Receptor_(biochemistry), which affect the chemical and electrical properties of neurons in dozens of ways.

>>4667861
This is dogma. There's little reason to believe that exogenous neuroactive chemicals are strictly bad.

The brain is finely tuned for an environment that isn't necessarily the environment we live in. Remember that the demands of civilization are very young, but our neuroanatomy is very old.

>>4665048
You again?

Leave /sci/ with your one-paper "conclusive proof" garbage. You did this last time with that Noetic trash science, too.

Many of the faculty I've spoken to detest instructing undergraduates. Understandably so - I'd hate teaching a bunch of snotnose brats some basic shit that I've had down for ~10 years if I was trying to publish and submit a grant a month.

>>4663353
>>4663354
OP was not criticizing B.S. level Biology, he was criticizing the field. Bachelor's level Biology classes are not an indication of the field's rigor. On the other hand, I had no trouble understanding these papers as an undergrad (nor did I have any trouble speaking to the authors about their work).

>>4663352
The brain is a biological system subject to the same constraints as other biological systems.

>>4663343
That work got Wang a tenured faculty position at Yale. http://wang.medicine.yale.edu/

He even used his models to predict the concentration of NMDA receptor in frontal cortex relative to other cortical areas.

>>4663311
Oh here's a good one. Just pop it open and read - If you're just entering year 2 of any physical science or engineering undergrad program, you shouldn't have any issue with it. It's about using statistical mechanics and information theory to talk about the efficiency of biological systems at detecting chemically (and behaviorally) relevant signals.

http://www.princeton.edu/~wbialek/our_papers/mora+bialek_10a.pdf

>>4663311
Can I cop out and just link you to a paper where a Physics Ph.D talks about describing LIP neurons with a diffusion model? It'd be a pointless exercise to rewrite his discussion section.

http://www.sciencedirect.com/science/article/pii/S0896627302010929

>excerpt
The attractor model is to be contrasted with accumulator or counter models commonly used in psychology [Ratcliff 1978], [Ratcliff et al. 1999] and [Luce 1986]. A similar model was proposed to describe the neural process in LIP for the visual motion discrimination experiment (Ditterich et al., 2001). According to this “integrate-and-decide” model, an LIP neuron selective to the left direction (L) integrates the input sL plus random noise, in the sense of mathematical calculus, so that its firing activity rL is described by drL/dt = sL + noise. Similarly, a competing neuron selective to the right direction (R) is described by drR/dt = sR + noise. Consequently, the difference X = rL - rR obeys the equation

X(t) undergoes a random walk in real time (a diffusion process), biased by a constant ramp (sL - sR)t (the rate of accumulation (sL - sR) increases with the signal strength, sL = sR for zero strength). When X(t) reaches a prescribed threshold ? (respectively -?), decision is reached for choice L (respectively choice R), and the reaction time RT = t is registered. The model for the saccade generation system proposed by Carpenter [Carpenter 1981], [Carpenter and Williams 1995] and [Reddi and Carpenter 2000] is similar in spirit, with an emphasis on the assumption that the rate of linear accumulation fluctuates randomly from trial to trial.

>>4663294
What kind of tests did you do, one-way ANOVA? Have you tried pairwise t-tests? All of this can be done in microsoft excel or a statistical programming language in under 10 minutes.

Also, what do you mean by "increasingly difficult?" Was every group performing just as bad as one another? You could say that the difficulty of your task kept your subjects on the floor (i.e., most of your data points have the lowest possible measurement on whatever dimension you're measuring). If all your measures were on the floor, it'd be very difficult to observe mean differences or differences in variability between groups without enormous sample sizes.

I expect second authorship, btw.

Biological systems are still under the same constraints as any other physical system, and can be described using the same physical models. Take the work of William Bialek as an example of that. Biology is as much a hard science as the variables you're measuring and the models you're plugging those variables into.

Come now, give us more information than that.

What's your discipline/what kind of measures were you using? What kind of variability did you observe relative to the range of your data? What were your sample sizes?

What kind of career do you want? Are you certain that a bachelor's in pure mathematics is ideal for you?

This is neither an experiment nor is it psychology.

Programming.

The likelihood that someone read this thread and was convinced by IQ fundie's arguments and GCP's data upsets me. A lot.

If you were convinced and are still reading this thread, just look at who's funding this garbage science:
http://www.noetic.org/research/overview/

We have no reason to believe that this data isn't fake.

go to bed

First of all, shame on OP for saying that such a correlational dataset "proves" anything.

Secondly, I'm skeptical that their low p-values are due to anything but confirmatory hypothesis testing in selecting events from the news based on observed z-scores. To do this PROPERLY, they would need to have a lot of people following the news in many different regions in order to get something that resembles a "complete" set of events that meet their "rigorously defined" criteria. That's a whole lot of news which would require a whole lot of people.

So, what kind of funding does this project get? Can they actually support having people following the news like this?

Moreover, have they used the correlational data they've been collecting to conduct any controlled experiments?

4) Talk to faculty and career advisors. Get an internship in a biotech/pharm/chemical engineering company at some point during your undergrad or your graduate studies. Make sure you do your doctoral thesis on something that has an obvious application.

5) I'm not familiar with any of these texts. I'd say cross reference those with whatever is in your university's library. When I self study at home, I tend to just use khan and online problem sets. Also, do yourself a favor and get some computer programming experience on your transcript - that skill got me an offer at a great school (Vanderbilt), and made 100% of my undergraduate research possible.

6) Again, talk to people. Talk to your professors and look up research scholarships your school offers. Talk to graduate students in your lab. Every lab and institution differs on how authorship is handled. If you maximize the time you spend in labs, then you should be able to maximize your productivity. I should mention that I was admitted to my program with my name on 0 published papers, but I've worked on four different projects (two of which were on my CV) and have presented them at conferences.

7) I took only the regular GREs. DO NOT take a subject test unless you're absolutely positive you can ace it - a high score (750+) may help you, but a low score (<700) WILL hurt you.

I was admitted in a high-ranked (ivy league) fully-funded Neuroscience Ph.D program for the upcoming fall, for what that's worth.

Contrary to what the previous poster says, I think its great that you're preparing now. If I had known all that I know now as an incoming freshman, I'd be at a slightly better school.

1) Graduate schools DO look at what's on your transcript. Something prettier, like honors, can only help you. It matters a lot less than research experience, though. Honors courses may help you get into a lab as an undergraduate or get scholarships to move along all that beautiful research experience, though.

2) Google scholar, and face time with faculty. Talk to everyone you can and find out more information. Be a little exploratory with courses you take - breadth is better than depth during your undergrad.

3) I don't know the specifics of chemistry, someone else is going to be more helpful here. Don't worry so much about making "non-trivial" contributions during your undergrad (its hard to guage that, anyhow) - just try to do original research over which you have creative ownership. First author on a manuscript that gets rejected several times is better than say, preparing buffers for a nature paper that you don't have authorship on (for getting into graduate school).

You can get by in (a lot of) biology without math, but that's not a very employable undergrad degree. Neither is a master's in Biology.

If you don't want to go to graduate or medical school, spend some time on khanacademy. If you want to go to graduate or medical school, spend some time on khanacademy.