Click HERE to see this thread with all of the graphics, features, and links.

We hear and read about studies and statistics used in those studies. But how objective is science, really?





Discuss statistics and studies: specifically, the methods used in both. Express your thoughts regarding the thread subject. This is not as broad of a topic as it seems.

No statistics isn't science, statistics is a methodological technique. Like I'm not sure I get the question. Are rulers science? Uh, no, dadude they're rulers...

Anyway onto the topic.

Competing schools of thought make it hard to describe statistics as generally objective or subjective. I like Andrew Gelman's take on subjective Bayes "if you could really construct a subjective prior you believe in, why not just look at the data and write down your subjective posterior?" (and he's essentially a Bayesian so no bias there). Simpler statistical methods have no room for subjectivity in their use by the scientist (there are no judgement calls in a chi-squared test or t-test) although you could argue they're ultimately based on subjective or arbitrary assumptions.

Choice of school, choice of test, and choice of data (as well as method of generating or collecting it) are all subjective factors that need to be considered when you try to interpret statistics.

Basically my view is this: Numbers don't lie but it's very easy to fool yourself.

I'm glad you hit Bayes... man, if only there weren't a dozen things I wanted to write for this forum.

I've got lots to say here... give me a couple of days.

Originally posted by Oliver North
I'm glad you hit Bayes... man, if only there weren't a dozen things I wanted to write for this forum.

I have a copy of Kruschke's book "Doing Bayesian Data Analysis" and I've been slowly going through it (I clearly need a better calculus background). I don't usually plug things but if you're looking for an introduction to Bayes rather than a sermon I'd really recommend it.

Originally posted by Oliver North
I'm glad you hit Bayes... man, if only there weren't a dozen things I wanted to write for this forum.

I've got lots to say here... give me a couple of days.


You can have 3.

Originally posted by Symmetric Chaos
No statistics isn't science, statistics is a methodological technique. Like I'm not sure I get the question. Are rulers science? Uh, no, dadude they're rulers...

The Science of Statistics: is it really "Science"

As in "originating from".

Or this:



For instance, a study that analyzed a sample and drew conclusions. The "science of statistics" portion would be all the science done involving statistics. As you call it: "methodologies."