Quote:
Originally Posted by daz
I agree too that math and music is related, but curious if it speaks for our preferred genre of music?
*intrigued*
|
Are you agreeing with me? .. Because I never wrote that I assumed that taste for music will go along with taste for maths. I think I have heard something along those lines, but I'd like to read an article before I make up my mind.
My comment was regarding this survey - I have an inlking that these questions will in fact
not give very reliable data on how music and maths correlate (or don't).
.. So if it was my survey, I'd rewrite the questions to account for more variables. The hidden ones!
I'll add (jokingly
) that if music and maths really did correlate positively, more people would understand the last two posts I wrote. I've been writing about statistics after all.
EDIT :
For the OP :
I'll give a hint as to how to improve this survey a bit. First of all, I think the approach might be a little off to start with. Try and think like this; what is it you want to show with the data for your survey? What are you investigating? How do you show it? I suggest coming up with a hypothesis.
Hypothesis : Musical and math skills correlate (means that skill at music and maths relate to eachother, either positively (hand in hand) or negatively (bad at one makes the other better))
You should also have an alternative hypothesis, like this :
Alternative hypothesis : Musical and math skills do not correlate (means your musical skills don't matter when it comes to maths)
The point is that evidence in favour of one will disprove the other one. Hypotheses like these are easy to test and get a good answer out of. Before you can start testing, you need to gather data to back up your hypotheses .. And then you need to make a survey that gives you some kind of good data. For example :
How many instruments do you play?
How many years have you played instrument(s)?
On a rating where 1 is lowest and 10 is highest, where do you rank yourself at math skills?
Putting some thought into questions is important. I mean, how are you gonna use an answer like "bad" in statistics? Nothing's impossible, but designing better questions may remove the problem completely.
Finally, you should watch out for hidden variables. Sex and age are accounted for (very good!), but there could be others. For example, as I mentioned, people who play more instruments may come from more wealthy families (because they can afford them) and if this really is a trend, that could influence your data (if being rich or poor influences skill with maths or music). To account for something like that, you need to make up more questions.
How would you describe your family's economic situation as you grew up?
You could make a sort of rating, f.ex - 1 = Poor, 2 = lower middle class, 3 = Upper Middle Class, 4 = Higher middle class, 5 = Rich.
I know that a question like that might be getting a bit intimate for people answering on a public forum, but just some suggestions to get your thoughts spinning. From the posts here, it seems to me like the question about genres for example is not gonna help you much in a statistical sense. I mean, maybe you could say something like "people who listen to classical music is smarter than people who listen to rock", but I don't think I'd believe in it and frankly .. you should have a pretty large sample size for anything like that.
Well, that's it for statistics 101 for now, but feel free to ask questions.
