Member Since: May 11, 2013
Answers: 1
Last Update: May 11, 2013
Visitors: 205
|
| |
Hi one of my final pieces of work for the year was a group project on a report, I somehow got lumped with the results section of a study I didn't even want to do so I'm at a loss for how to do it, if anyones good at statistics please help :)
This is the basic study:
We have a sample of 100 (50 males, 50 females) and are looking at oppinions on subcultures. To do this we produced our own questionnaire with 5 questions on a 1-10 Likert Scale for each mentioned subculture
e.g.
1). how approachable is a person from this subculture
2). Can you communicate with a person from this subculture
3). How deviant are people from this subculture
4). Would you feel comfortable with someone from this subculture
5). Do you feel this subculture is misrepresented?
So I know I have Interval data and we had three hypothesis
A). Goths are less approachable than Skaters and geeks
(I think this is a pearsons correlation)
B). there may or may not be a difference in oppinion between goths and emos
c).There will be a difference in the views of the subculture null hypothesis
If anyone could tell me what tests to use that would be super great! I can work out how to do them myself but all my reading around has left me stumped. Thanks so much for your help! (link)
|
Well for A you can only have two samples unless skaters and geeks are in the same group but that may create a bias. You do the same as the next equation down but do not multiply by two. You use z if standard deviations for the average of the popluation is known or you use the t chart if you only the standard deviation for the sample
B you the Null hypth. would be there is no diverence and and the Alternative hypth. would be there is a difference. You would used the formula
X1-X2/ (sqaure root of (V1/n1)+(V2/n2))
V = variance
N = number per sample
1 is sample 1
2 is sample 2
The you look up the p then multiply by 2.
C. X - average / standard deviation / sqaure root of n
then look up the p value and multiply by 2, Z test unless you do not know the standard deviation for the average of the poplation but you do for your sample.
|
|