t-Distribution Example Problem and Solution
Problem:
1. A teacher believes that students in her class study an average of 8 hours per week.
A random sample of 10 students shows a mean study time of 9.2 hours with a standard deviation of
1.8 hours.
At a 5% significance level, can we conclude that students study more than 8 hours per week?
Given Data:
Sample mean (X_bar) = 9.2
Population mean (mu_0) = 8
Sample standard deviation (s) = 1.8
Sample size (n) = 10
Significance level (alpha) = 0.05
Step 1: Hypotheses
H0: mu = 8
H1: mu > 8 (Right-tailed test)
Step 2: Test Statistic
= (9.2 - 8) / (1.8 / sqrt(10))
= 1.2 / 0.569 = 2.11
Step 3: Degrees of Freedom
df = n - 1 = 9
Step 4: Critical t-value
At alpha = 0.05 and df = 9 (right-tailed):
t_critical = 1.833
Step 5: Decision
Since t_calculated (2.11) > t_critical (1.833), we reject H0.
Conclusion:
There is significant evidence at the 5% level that students study more than 8 hours per week on
average.
Decision Rules for the t-Test
We compare the calculated t-value (from data) with the critical t-value (from table).
Case 1: t calculated > t critical
Meaning:
The test statistic lies in the rejection region (far in the tail).
Decision:
Reject the null hypothesis H0
Interpretation:
There is significant evidence to support the alternative hypothesis H1.
Case 2: t calculated = t critical
Meaning:
The calculated t-value is exactly at the boundary of the rejection region.
Decision:
It’s a borderline case — often we reject H0(since it meets or exceeds the threshold).
But some statisticians consider it not significant unless t > t critical .
Interpretation:
Result is just significant (right on the edge).
Case 3: t calculated < t critical
Meaning:
The test statistic falls inside the acceptance region (not far from mean).
Decision:
Fail to reject H0
Interpretation:
There is not enough evidence to support H1.
We conclude that any observed difference could be due to chance.
2. A machine is supposed to fill packets that weigh 50.0 grams on average. A quality inspector
takes a random sample of 12 packets, and the sample shows a mean weight of 49.0 grams with
a standard deviation of 1.5 grams. At the 1% significance level, test whether the machine is filling
packets with a mean weight different from 50.0 grams. Use a two-tailed t-test.
Solution
Given Data:
Sample mean (x̄) = 49.0 g
Population mean (μ₀) = 50.0 g
Sample standard deviation (s) = 1.5 g
Sample size (n) = 12
Significance level (α) = 0.01
Step 1: State the Hypotheses
Null hypothesis (H₀): μ = 50.0 g
Alternative hypothesis (H₁): μ ≠ 50.0 g
This is a two-tailed test because we are checking for any difference (either less or more) from 50.0 g.
Step 2: Calculate the Test Statistic
Formula: t = (x̄ - μ₀) / (s / √n)
Substitute the values:
t = (49.0 - 50.0) / (1.5 / √12)
√12 = 3.4641
s / √n = 1.5 / 3.4641 = 0.4330
t = -1.0 / 0.4330 = -2.309
Thus, the calculated t-value is approximately -2.309.
Step 3: Degrees of Freedom
df = n - 1 = 12 - 1 = 11
Step 4: Critical t-value
For a two-tailed test at a 1% significance level with df = 11:
t₍critical₎ = ±3.106
Step 5: Decision Rule
Reject H₀ if the calculated t-value lies outside the range (-3.106, +3.106).
Since t₍calculated₎ = -2.309 falls within this range, we fail to reject H₀.
Step 6: Interpretation
The test statistic (t = -2.309) does not fall in the rejection region. Hence, there is not enough
evidence at the 1% significance level to conclude that the machine’s mean fill weight differs from
50.0 g. However, if we consider a 5% significance level, the approximate p-value (0.04) would be
less than 0.05, suggesting that the difference would be significant at that level.
Conclusion
At the 1% significance level, we conclude that there is no significant difference between the actual
mean fill weight and the claimed mean of 50.0 grams. The machine appears to be operating within
acceptable limits.
3. A manufacturer claims their LED bulbs last 1200 hours on average. A consumer group tests a
random sample of 15 bulbs and finds a sample mean lifetime of 1160 hours with a sample standard
deviation of 80 hours. At the 5% significance level, test whether the true mean lifetime is different
from 1200 hours. Use a two-tailed t-test.
Solution
Given Data:
Sample mean (x̄) = 1160 hours
Population mean (μ₀) = 1200 hours
Sample standard deviation (s) = 80 hours
Sample size (n) = 15
Significance level (α) = 0.05
Step 1: State the Hypotheses
Null hypothesis (H₀): μ = 1200
Alternative hypothesis (H₁): μ ≠ 1200
This is a two-tailed test because we are checking for any difference (either less or more) from 1200
hours.
Step 2: Calculate the Test Statistic
Formula: t = (x̄ - μ₀) / (s / √n)
Substitute the values:
t = (1160 - 1200) / (80 / √15)
√15 = 3.87298
s / √n = 80 / 3.87298 = 20.6579
t = -40 / 20.6579 = -1.936
Thus, the calculated t-value is approximately -1.936.
Step 3: Degrees of Freedom
df = n - 1 = 15 - 1 = 14
Step 4: Critical t-value
For a two-tailed test at a 5% significance level with df = 14:
t₍critical₎ = ±2.145
Step 5: Decision Rule
Reject H₀ if the calculated t-value lies outside the range (-2.145, +2.145).
Since t₍calculated₎ = -1.936 falls within this range, we fail to reject H₀.
Step 6: Interpretation
The test statistic (t = -1.936) does not fall in the rejection region. Hence, there is not enough
evidence at the 5% significance level to conclude that the mean lifetime of the bulbs differs from
1200 hours. The approximate p-value is 0.074, which is greater than 0.05, confirming our decision.
Conclusion
At the 5% significance level, we conclude that there is no significant difference between the actual
mean lifetime and the claimed mean of 1200 hours. The bulbs appear to be performing as expected.