Z-score or T-score?? *statistics help plz*

[hello again 10+ year member]

I am doing a statistics project and I am basically comparing the GPAs of Freshman/sophomore college students VS. juniors/seniors. When doing my hypotheses testing and comparing the two, am i suppose to use normal model with z-scores? Or will I use T-model? Im not sure when to use each.. please let me know statistics experts!

Click to read more...

 

[synikalt 10+ year member]

Ok...seriously do you really think you could bring your math problems to a Car audio forum? statistics is hard..since i took it last semester i will be gladly to help.

I havn't open this book for about a year now..but im pretty sure you only use the T-model when your distribution is not normal and you have less than 30 samples. so then you t-model sample may be skewed, and you will use some formula to subract 1 from the sample to look up the according T-score value.

hope this helps..w/e..read ur book

 

[hello again 10+ year member]

hey man thanks a lot for the help. I was flipping through the book and I think you're right, I need to use T-score since I am going to be talking about averages of the GPAs which I am going to compare right? Z-score is more about working with proportions correct?

oh, and about the math work on caraudio forum, i posted this same question on a "school help" forum about an hour ago and still not a single reply. From my experience in random threads, I learned caraudios forum has so many ppl checking it so i should get a reply by at least one person //content.invisioncic.com/y282845/emoticons/biggrin.gif.d71a5d36fcbab170f2364c9f2e3946cb.gif either that or theres a lot of smart ppl on this forum.

thanks again

 

[synikalt 10+ year member]

yea..actually..u have the book...u could know more about this than me. i havn't seen the book in a long time. but T-score and Z-score acount for the same thing. the Z score equals the X (sample size) minus the mean devided by the stanard deveation. T-score account for the same thing but you look it up differently. theres formuls you have to abide by to find the correct answers. proportion thing you might be talking about sounds familiar. and the z-score has to do with it as well. but idk...its vauge to me. ask your teacher

 

[hello again 10+ year member]

yea i wasnt too sure about the t-score and z-score part but it hink thats right.. but basically i could still use the z-score for this but it just wouldnt be as accurate right? thanks anyway bro have a good one

 

[jmanpc 5,000+ posts]

i forgot all that shit after i took the test.

 

[faulkton 5,000+ posts]

use a z score.

 

[Nukedsodapop 5,000+ posts]

if you know the standard deviation, then use z, if standard deviation is unknown, then you use t. Thats how I remembered it, but I could be wrong

 

[Nukedsodapop 5,000+ posts]

and I think the data would be kinda biased towards the GPA of the college students because in high school juniors and seniors dont really give a crap but to go to college you actually have to care enough to go, but the work is harder so who knows

 

[PV Audio 5,000+ posts]

This is a two sample t distribution. Actually, before I say that, what are your hypotheses?

 

[Nukedsodapop 5,000+ posts]

probably

H0= H.S. GPA = College GPA

Ha= H.S. GPA Not equal to College GPA

 

[PV Audio 5,000+ posts]

Yeah I figured, but it all depends on what he's actually trying to do. What he said is that he's basically comparing the differences in height between blacks and latinos, but there's too much out there to make a clear call. For all we know, he could have some set proportion that we don't know about, or even better, the constant rates at which the GPAs change could use a little poisson distribution.

 

[lilmaniac2 5,000+ posts]

had stats last semester, got a A

i believe uve already got the correct answer so i just wanted to brag about making my A... hahaha gotta love a cheat sheet

 

[hello again 10+ year member]

if you know the standard deviation, then use z, if standard deviation is unknown, then you use t. Thats how I remembered it, but I could be wrong

i think thats wrong.. cant the standard deviation be found as long as you have the data?

 

[hello again 10+ year member]

probably
H0= H.S. GPA = College GPA

Ha= H.S. GPA Not equal to College GPA

Nooo i think you read it wrong. Im comparing college students in general. this is the hypotheses

Ho: College freshman/sophomore = college junior/senior

Ha: College freshmen/sophomore NOT EQUAL college junior/senior

i also plan to use a histogram and box plots for my descriptive statistics and then show work for doing the t-score two-sample t-test thing as my inferential statistics. Here is the data I collected by doing surveys

FRESHMAN/SOPHOMORE GPA's

3.5

2.5

3

3.4

3.75

2.5

3.8

2.4

3.23

3.8

4

3.5

2

4

2.5

3.3

3.87

3.8

1.8

3.7

4

3.4

3.75

3.25

3.2

3.67

4

2.7

JUNIOR/SENIOR GPAs

4

3.93

2.8

3.93

3.75

4

3.7

3.75

4

3.92

2.72

4

3.5

4

4

3.75

2.75

3.5

3.7

4

3.4

now its time to start punching data in my calculator and do some charts on excel. any advice and tips???