Paired samples t-test worksheet
KEY
You want to investigate if feeling hungry makes you buy more snack food at the grocery store. To test your hypothesis, you recruit 10 friends to help. You ask them to report the number of snack food items they buy at the store when they are not hungry as their baseline number. Then, you ask them to tell you the number of snack food items they buy the next time they go grocery shopping but haven’t eaten for at least 4 hours before the shopping trip. Now you have two measures for each of your friends. Conduct a paired samples t-test
to determine if people buy more snack food when they are hungry versus when they are not. Test the hypothesis
with a one-tailed test at α
= .05
. M
D
= 4.20
SS
D
= 48.80
1. Calculate the paired samples t-test and report whether you would reject or fail to reject the null hypothesis. Show all your work. s
D
= √
SS
D
df
= √
48.80
10
−
1
= √
48.80
9
= √
5.42
= 2.33
s
MD
= s
D
√
n
= 2.33
√
10
= 2.33
3.16
= 0.74
t = M
D
−
μ
D
s
MD
= 4.20
−
0
0.74
= 5.68
df = 9
t
crit
= +1.833 Reject null because 5.68 > 1.83 2. Write up the results in APA style. People buy significantly more snack food when they are hungry versus when they are not, t
(9) = 5.68, p
< .05.
3. How big was the effect size for this study? Calculate either Cohen’s d or r
2
and describe your result in words. Cohen’s d = 4.20/2.33 = 1.80; People bought 1.80 SDs more snack food when they are hungry compared to when they are not. r
2
= (5.68)
2
/(5.68)
2
+ 9 = 32.26/(32.26+9) = 0.78; Being hungry accounted for 78% of the variance in amount of snack food bought.
