T-test
One Sample t-test using dummy (two- tail):
Problem Statement- To determine that there is a significant difference between estimated
average hours of TV watched per week (28.5) and calculated hours at alpha=5%.
Dumm
Hours y
25.7 0
38.5 0
29.3 0
25.1
30.6
34.6
30
39
33.7
31.6
25.9
34.4
26.9
23
31.1
29.3
34.5
31.2
33.2
30.2
36.5
37.5
27.6
24.6
23.9
27
29.5
30
29.6
HYPOTHESIS TESTING:
Null Hypothesis: There is no significant difference between estimated average hours of TV
watched per week (28.5) and calculated hours
Alternate Hypothesis: There is a significant difference between estimated average hours of
TV watched per week (28.5) and calculated hours
H0 = µ = 28.5
H1 = µ ≠ 28.5
Output:
t-Test: Two-Sample Assuming Unequal
Variances
Dumm
Hours y
Mean 30.48276 0
Variance 19.13362 0
Observations 29 3
Hypothesized Mean Difference 28.5
df 28
t Stat 2.441014
P(T<=t) one-tail 0.010613
t Critical one-tail 1.701131
P(T<=t) two-tail 0.021226
t Critical two-tail 2.048407
DECISION RULE:
If t-stat is greater than t-critical, reject null hypothesis.
If p(t) is less than α, reject Null hypothesis.
INFERENCE:
Since t Stat(2.44) is greater than t Critical(2.04), reject Null hypothesis.
Since P(0.02) is less than α(0.05), reject Null hypothesis.
CONCLUSION:
There is a significant difference between estimated average hours of TV watched per week
(28.5) and calculated hours