How Accessible Are Data Sets from Peer-Reviewed Journals?

[27Thousand]

I asked this in the mathematics section, but they told me it was probably best to ask it in a different PF section. So I'm interested in getting a hold of peer-review journal articles' "original data sets".

I'm curious because I want to play around with data for fun on some statistics software I'm teaching myself, plus I want to create equations to describe/predict things I find interesting in peer-review journals. I've been looking at peer-review journals for fun since I was in high school. Not that any of us are Isaac Newton, but he was into making equations to describe/predict things. It helped him come up with new hypotheses to test, see patterns between already existing ideas to create broader perspective principles, and to see which ideas may have looked the same but which were extremely different.

How often do they let others see their original data sets? Do people just email the authors of a peer-review article to get it from them?

 

[Monique]

Original data sets are often published as supplementary material. Before you start to play around with data, you will first need to become intimately aware of how the data was acquired: you need to be aware of the limitations of the data.

Also note: doing multiple tests can be informative as a test case, but it is very problematic when you want to draw a real conclusion.

 

[27Thousand]

One of the main reasons is because I want to learn how to read women, when flirting. Yes I know that sounds nerdy, but an equation may be a new and useful way of dealing with things.

Then another area of interest, I'm intrigued by research on how genetics affects behavior, like with the twins separated at birth studies (IQ and other aspects of personality). I also think it's interesting how neurologists say more successful neural pathways become strong and less successful weaker. I want to look at data at the macro level of behavior and visualize in my mind how these two concepts of environment/genetics come together.

Original data sets are often published as supplementary material. Before you start to play around with data, you will first need to become intimately aware of how the data was acquired: you need to be aware of the limitations of the data.

Also note: doing multiple tests can be informative as a test case, but it is very problematic when you want to draw a real conclusion.

I agree, I'll have to keep in mind the context of the situation how the study was conducted.

This is what I'm thinking about to keep what you're saying in mind. As you know statisticians have methods of creating weak equations in the social sciences which look something like this

As you many here probably know, if you want more than one X independent variable to predict the Y dependent variable you can use multiple regression and create an equation like:

y^ = b0 + b1x1 + b2x2 + ...

So what I'm thinking of is just like you said, you have to keep in mind the context of the situation how they collected the data. Perhaps, maybe then I could somehow discover a BnXn variable to add. Actually that's one of the reasons I'm interested in this, so I can find out how data differs from study to study, any universal patterns regardless, plus any universal patters of how it differs from context to context (even the context may have patterns to it).

I may even possibly make a discovery. Then in addition to that add other mathematical concepts to the equation so it's more original than just multiple regression or any of the other ways statisticians make data fit an equation. Even if someone's already thought of any equations/patterns I may discover, I just think it would be an interesting experience to discover things on my own.

 

[Redbelly98]

How often do researchers from peer-review journals let others see their data sets?

It probably depends on who asks for the data. They could be willing to share data with colleagues and former collaborators, but not willing to share with a total stranger who is not an active, professional researcher.

 

[mgb_phys]

Depends on the source of the data - anything obtained from public facilities (Nasa, national telescopes) is available after a proprietry period (often 1-3 years)

 

[Evo]

Sorry, but based on the fact that the OP is only interested in doing this for dating, thread closed.