Testing for two means

[kstone]

I am doing practice questions for hypothesis testing, for two means, n is large.

I know the test stat for two means.

H0: Mu1 - Mu2 = 0, Ha: Mu1 - Mu 2 > 0

The data is:
Male: X: 6.8, s: 6.6, n: 291
Female: X: 2.4, s: 4.5, n: 519

After doing the problem, the answer I received was 10.23, I want to check whether this is correct as it seems too large?

Thanks!

 

[TheEcologist]

Hi Kstone,

Why don't you show us your calculation steps, so we can see if you're doing it right.

TE

 

[kstone]

Male: X: 6.8, s: 6.6, n: 291
Female: X: 2.4, s: 4.5, n: 519

Using the test statistic formula for two means:

6.8 - 2.4 - 0
/
6.6^2 / 291 + 4.5^2 / 519 (square root answer)

=

4.4 / 0.434

= 10.23

It seems too large to me, as all times I've practised other questions I haven't gotten an answer above 4.

 

[TheEcologist]

Hi KStone,

Your calculation steps seem good to me (you are correcting for unequal variances and unequal sample size) .

(6.8-2.4)/sqrt(({6.6^2}/291)+({4.5^2}/519)) = 10.13 (small difference here with your answer, rounding errors?)

So as far as I can tell you doing it correctly.

Cheers,

TE

 

[kstone]

Excellent, good to know I'm not way off.. thank you!

 

[Kidara]

Hi KStone..

(small difference here with your answer, rounding errors?)

I agree. Looks good but it's likely the early rounding is affecting your result. Try to avoid rounding numbers until the very end, otherwise you may damage the strength and meaningfulness of your results.

Luck,
Kidara.