Education Jobs

[Siggie]

Re: Education Jobs

Oh and a correction - Correlation does not equal causation is misleading. It would be more accurately stated as correlation does not necessarily equal causation.

There are 3 things required for causal inference: 1) covariation (correlation) 2) temporal precedence and 3) freedom from confounding (e.g. by random assignment).

It's not about what the statistical analysis was, but rather about methodology. If the methodology warrants a causal inference, then it's appropriate even if the analysis conducted was, for example, computing a Pearson's R or something.

[Mos]

Re: Education Jobs

Violating that principle is bad science and will be called out in peer reviews - the same biases exist in all fields of science so what is your point? Math is not just the basis for physics and chemistry it is the bases for all other fields to including social sciences. Just because you disagree with the facts that are found in lets say psychology or sociology does not make these fields bad science.

So what I'm reading from that is that you did an undergraduate major in applied mathematics and worked as a research assistant or something on some kind of research. Hardly enough to claim any sort of "expertise" in any scientific field.

Let me try being less critical and more instructive ... Science is a method. It isn't a subject. Science is what you apply to a field defined by a theme in the type of research questions that are asked. E.g. Political science dealing with research questions having to do with politics and political beliefs and the like. Science is science is science. Whether the question "How do we perceive color?" is asked by a neurologist, psychologist, or physicist is not relevant. If we apply the scientific method in attempting to answer the question and test related hypotheses, we are doing science.
Your disappointment stems from the SUBJECT of the scientific inquiry. It is more difficult to study people because they vary much more than white rats, which vary much more than samples of a given element. It's harder to do and to control for extraneous, potentially confounding variables, but we do our best and we still apply the exact same methods.
There are programs in these fields which are a little more on the humanities side than the social science side perhaps and certainly quality varies greatly among programs at different institutions. In some classes the way in which the information that is communicated was learned isn't emphasized and that's too bad in my opinion. When I teach, I don't create that distance between the acquired knowledge and the application of scientific inquiry which led us there. I take a more scientific, methdological emphasis, but not everyone does. Also, fyi, I have friends from these "hard science" fields who have said to me that I (background in Psych) have more research methods and stats training than they do, because they are trained to apply in their narrower, much less variable fields, and they learn only the limited methodologies/analyses that typically use. Whereas, I seem to have expertise with methodology and stats that is more easily applied across disciplines and in "real world application." So, it's really specifically about how much relevant "how to do science" training one receives and not the field in which they receive it. To make a blanket statement that Poli Sci is not a science is...well, ignorant.

Oh and a correction - Correlation does not equal causation is misleading. It would be more accurately stated as correlation does not necessarily equal causation.

There are 3 things required for causal inference: 1) covariation (correlation) 2) temporal precedence and 3) freedom from confounding (e.g. by random assignment).

It's not about what the statistical analysis was, but rather about methodology. If the methodology warrants a causal inference, then it's appropriate even if the analysis conducted was, for example, computing a Pearson's R or something.

Of course it's about methodology. If you don't have things such as random assignment and random selection, the power of your inference is greatly weakened. Researchers fall victim to all sorts of biases and sometimes fail to ensure such randomness that must be present in the study. To be able to infer something about a general population is a very strong thing. You need a lot supporting you to make such an inference and you need to ensure your readers about randomness and how you've minimized certain biases. That's why studies in sociology/psychology can be so tricky and very hard to accurately assume something about the general pop in these fields and you need a lot of evidence in your data and methodology to prove the power and legitimacy of your inference.

[Siggie]

Re: Education Jobs

Of course it's about methodology. If you don't have things such as random assignment and random selection, the power of your inference is greatly weakened. Researchers fall victim to all sorts of biases and sometimes fail to ensure such randomness that must be present in the study. To be able to infer something about a general population is a very strong thing. You need a lot supporting you to make such an inference and you need to ensure your readers about randomness and how you've minimized certain biases. That's why studies in sociology/psychology can be so tricky and very hard to accurately assume something about the general pop in these fields and you need a lot of evidence in your data and methodology to prove the power and legitimacy of your inference.

I said causal inference specifically. That may not be the case at all; it depends entirely on the research question(s). And the population to which you wish to generalize the findings depends on the research question as well. We're not always interested in universally true phenomena exclusively. Sometimes we want to know about a smaller population; e.g. alzheimer's patients.
There are also lots of universal findings that have come from psychology too (I'm speaking about Psych because it is my field). E.g. how we recognize/process faces, that it's faster to relearn than learn first time, cue-dependent learning, inattentional blindness, developmental sequences, perceptual processes, etc. etc. etc.

Anyway, what's the bottom line? You're saying it can be more difficult to do science in these fields, but that it's doable?

[KanadaHye]

Re: Education Jobs

Hit a nerve, eh?
You do know that science is about more than math right?

Not really. There is a word for science without math. I think it's called philosophy or religion. Every field of applied science involves math.

Let me reiterate. If you can't quantify or analyze your research with the use of statistics, your observations are merely subjective and aren't factual.

[Siggie]

Re: Education Jobs

Not really. There is a word for science without math. I think it's called philosophy or religion. Every field of applied science involves math.

Just to be clear, when you say "not really," you're disagreeing with the statement "science is about more than math"?

If so, then I see that you have limited your reading and comprehension to one sentence again and ignored the rest of my words where I said to do science you must use math, but math isn't the end all be all because the methodology is essential. Won't you ever tire of trying to build strawmen?

[KanadaHye]

Re: Education Jobs

Just to be clear, when you say "not really," you're disagreeing with the statement "science is about more than math"?

No, I'm saying that math is what separates science from science fiction.

[KanadaHye]

Re: Education Jobs

If so, then I see that you have limited your reading and comprehension to one sentence again and ignored the rest of my words where I said to do science you must use math, but math isn't the end all be all because the methodology is essential. Won't you ever tire of trying to build strawmen?

I find this funny because you're the one that attacked Mos and implied he didn't know anything about science when what he said was true.

Please, please, please don't tell high school students that math isn't integral to a future in the sciences.

[Siggie]

Re: Education Jobs

I find this funny because you're the one that attacked Mos and implied he didn't know anything about science when what he said was true.

Please, please, please don't tell high school students that math isn't integral to a future in the sciences.

There you go with the strawmen STILL. I said nothing of the kind.
You know what... you're coming in here mid-discussion and you're completely WRONG about what agreement we came to.

I said social sciences do use math because you need math to do science because of the probability/inference testing. But that it isn't just math for the sake of math because it's the methodology that's key. Learning calculus doesn't make one a good scientist if they don't know the first thing about how to apply the scientific method. If it was only about Math then we wouldn't have two fields called maths and sciences because they'd be one and the same. You do need math, yes, but math is but one piece. Don't misrepresent what I said and then disagree with that misrepresentation.

[KanadaHye]

Re: Education Jobs

Because PolSci has very little mathematical and concrete theory to it. It's often very subjective and not objective enough as other sciences. A field such as Finance or Chemistry have a much more objective and pure theory to it.

You don't know sh!t about science. I love when someone who's never taken much coursework in a subject states something they're totally ignorant about with such great confidence. It's a dangerous thing to think you know a lot more than you know. Be humbler and actually look into something before speaking on it.

PolSci has very little mathematical and concrete theory? Biology is very mathematical then? All sciences use probability theory and statistical theory. If you want to get into the equations for the statistical tests we use to avoid spurious findings and errors, I have a whole book of equations for you. Many from calculus. It will satisfy your "math" requirement.

There you go with the strawmen STILL. I said nothing of the kind.
You know what... you're coming in here mid-discussion and you're completely WRONG about what agreement we came to.

I said social sciences do use math because you need math to do science because of the probability/inference testing. But that it isn't just math for the sake of math because it's the methodology that's key. Learning calculus doesn't make one a good scientist if they don't know the first thing about how to apply the scientific method. If it was only about Math then we wouldn't have two fields called maths and sciences because they'd be one and the same. You do need math, yes, but math is but one piece. Don't misrepresent what I said and then disagree with that misrepresentation.

i GiVe uP

[Siggie]

Re: Education Jobs

Exactly. I love that you quote me AND STILL GET IT WRONG.

"All sciences use probability theory and statistical theory. If you want to get into the equations for the statistical tests we use to avoid spurious findings and errors, I have a whole book of equations for you. Many from calculus. It will satisfy your "math" requirement."

The claim was that polisci is a lesser science because there's not enough math. I said that what defines a science is the utilization of the scientific method as a means to acquire knowledge. Therefore if a field is using that method, which requires application of mathematics (e.g. statistics, probability theory, modeling, etc.) then it's science.

Which part of that is confusing? I am saying that if you're doing science then the math comes with it.