Science Buddies: "Ask an Expert"

Hello. We would like your opinion on an idea of ours. We would like to research how IQ can affect your GPA (visa versa) in the class subjects english, math and science. We would appreciate any thoughts, whether they be positive or critisicm. Thank you very much.

Hello Andred,

Looking for a correlation between GPA and IQ sounds very interesting. However, there is one issue you may want to consider.

Where are you going to be getting your data? Do you have access to individual's GPAs and their IQs? I would imagine that there's quite a bit of privacy issues with learning peoples grades and their IQs. If you can learn this sort of information anonymously, I think it would be fine. Otherwise, you will need informed consent forms signed.

Almost all science fairs follow SRC guidelines. Here's a link to the webpage with the guidelines concerning projects that involve human subjects in any way:

http://www.sciserv.org/isef/teachers/rules/rules7.asp

Also, keep in mind the ethical implications of your work. Many people place too much stock in IQ scores relating directly to intelligence. In fact, there have been many studies demonstrating that IQ tests only measure one type of intelligence, and there are many, many different modes out there. You may consider looking at two different intelligence tests if you want to draw significant correlations. A Google search for "intelligence tests" will yield a number of results.

Good luck with your project. It certainly sounds like an interesting idea.

Sincerely,
Linh Pham

Hey Andred,

This is a good research topic, although the answer to the question is quite predictable. You could find a website about the relation. You could also find a website with an IQ test and tell a lot of people to take the test. Then, you could ask those people what their GPAs and IQ scores are, and keep it on a scatter plot.

Good luck
aznnerd666

When you depend on people to voluntarily provide you with information you introduce serious statistical bias that will put your results in doubt.

1) Self selection means that your sample chooses itself rather than you choosing your sample. Ideally, you want to randomize your sample by age, gender, and whatever other variables are relevant. This sample serves as a model of the total population. A self-selected sample is not a model of the general population, but of an unknown sub-population whose properties you know little about. Conclusions drawn from this sub-population may not be valid for the general population.

2) If you are depending on people accurately reporting their GPA or IQ, information which has a high emotional charge and can be socially stigmatizing, you run a high risk of receiving false information. This will also be related to the self-selection problem, in that perhaps only people who are not embarassed about their IQs or GPAs will form your sample.

There are also many unresolved questions about what GPA and IQ actually measure, as well as what bearing they have on people's lifestyles. There is no universally agreed upon definition of human intelligence, it means different things in different contexts, and slips through your mind's fingers like water. This is not a simple project at all, and you run a very high risk of angering a wide variety of people. Beware.