# Thread: paired t test problem, just check

1. ## paired t test problem, just check

this website gives a paired t test problem

http://www.texasoft.com/winkpair.html

it doesnt provide an answer so could someone please tell me if i did it right.

its basically about pizzas being compared to see if type a or type b is more preferred

null = both pizzas are equally favored i.e. null: A - B = 0
alt = 1 is more favored over the other i.e. alt = A - B > 0

using a 1 tailed t test

t = average difference/standard error of difference

sample size = 10
avg difference = 4.8
standard deviation = 4.45620
standard error = 1.4092

t = average difference/standard error of difference
t = 4.8/1.4092
t = 3.406

t-tabulated = 1.833 >> found using a t table with a confidence interval of 95%

since t - calculated > t - tabulated therefore reject null so yes there is a preference of 1 type over the other

...was i correct?

thanks

2. ## Re: paired t test problem, just check

Looks right to me.
Though you can install megastat or kaddstat addin for excel and using these packs you get a p value as well so it is easier to infer the answer that way.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts