2. ## Re: Test of Significance?

Is their a way to compare how much better my chances are if i had a 3.52 vs a 3.47? If the average is 3.5?

3. ## Re: Test of Significance?

Well we'll just ignore that the average is 3.5 and use logistic regression. Logistic regression is my advice. If you need help implementing it then let us know what stats packages you have available or are familiar with.

4. ## Re: Test of Significance?

I've never learned about logistic regression although I have learned about linear regression. I was hoping to implement something we have learned in the school year, so he would be impressed that I learned and im interested. I know about probability, and a bunch of tests (Ex: t-test, chi sqaured test, 2 sampled t test, etc...)

5. ## Re: Test of Significance?

Once again: if you don't do it proper then what is the point? I understand you probably haven't learned about it - but if you're actually hoping to increase your grade I doubt doing a simple test is going to accomplish your goal. I don't think any simple test will actually tell you what you want to know either. You could try to test if the mean acceptance rate for those with a GPA higher than 3.52 is better than the mean acceptance rate for those with a GPA of 3.47 or lower. But that really doesn't tell you what you want.

I can understand your dilemma in that you only know some simple tests. But I also understand where a teacher is coming from and if you don't do it proper then they might just think "Well you're lucky to have received the grade you did because this is horrible and misleading". Doing something that is beyond the scope of the course, though, might actually show the teacher that you're interested in the material.

Now like I said you probably should just talk to your teacher because realistically they probably won't change your grade and there isn't much you can do about it (that is if the course has already been completed - which it seems to me is the case).

6. ## Re: Test of Significance?

i think you would be better of spending your time taking another course to improve your gpa columbia. second session summer school is about to start if you are in the US.

7. ## Re: Test of Significance?

Hehehe...the guy's persistent...even when he doesn't know what he's doing. I gotta give him credit for that.

I agree with Dason though. If you don't do it properly, then what's the point? The answer will only be wrong, the prof WON'T be impressed (if anything, though I'd be touched that you took the effort, I'd be unimpressed cause you didn't do it correctly).

Further, even if you DO do it (and somehow it manages to be correct), are you able to defend it? Explain and justify your decisions? Present it to him with logical interpretations?

I can tell you're somewhat desperate to try to improve your grade. I also think the best thing to do is to just talk to your professor. Tell him what you're interested in doing and see if he/she will work with you. If not, you can take extra courses to improve your GPA.

8. ## Re: Test of Significance?

And even if you do the analysis there's no guarantee that you'll like what you find. It might turn out to be that that small of an increase in you GPA really doesn't make much of a difference to the probability of getting admitted. Which I'm assuming would not necessarily help your case?

9. ## Re: Test of Significance?

has anyone else realised that this post has almost 300 views (probably more by the time someone else reads this)??? stay tuned!

10. ## Re: Test of Significance?

Originally Posted by Dason
Once again: if you don't do it proper then what is the point? I understand you probably haven't learned about it - but if you're actually hoping to increase your grade I doubt doing a simple test is going to accomplish your goal. I don't think any simple test will actually tell you what you want to know either. You could try to test if the mean acceptance rate for those with a GPA higher than 3.52 is better than the mean acceptance rate for those with a GPA of 3.47 or lower. But that really doesn't tell you what you want.

I can understand your dilemma in that you only know some simple tests. But I also understand where a teacher is coming from and if you don't do it proper then they might just think "Well you're lucky to have received the grade you did because this is horrible and misleading". Doing something that is beyond the scope of the course, though, might actually show the teacher that you're interested in the material.

Now like I said you probably should just talk to your teacher because realistically they probably won't change your grade and there isn't much you can do about it (that is if the course has already been completed - which it seems to me is the case).
I agree I should do it the right way, im going to research more about logistic regression. As for my grade being changed, its worth a shot to do this and my teacher loves it when kids do this kind of stuff...

Any advice on how i can get started with that?

11. ## Re: Test of Significance?

Did you use any stats software in your class?

12. ## Re: Test of Significance?

We used a TI-84 calculator and all the programs that it carries for tests and such... but nothing intense..

13. ## Re: Test of Significance?

You would probably need something more sophisticated than a TI-84. I don't use Excel but it might be able to handle logistic regression. Otherwise there is R which is free and open source but has quite a bit of a learning curve.

Edit: The other option being choosing a data set and a question you actually learned how to answer in your course. The main problem right now is that you didn't cover using a continuous variable to predict a binary response.