A researcher tests whether the mean recovery time from surgery is different from the hospital's standard of 5 days. From a random sample of 36 patients, the sample mean is 5.8 days and the sample standard deviation is 1.8 days. The test statistic is:
- At = (5 - 5.8) / (1.8/36) ≈ -16
- Bt = (5.8 - 5) / (1.8/6) = 0.8 / 0.3 ≈ 2.67Correct
- Cz = (5.8 - 5) / (1.8/6) ≈ 2.67
- Dt = (5.8 - 5) / 1.8 ≈ 0.44
Why B is right
One-sample t-test: t = (x̄ - μ₀) / (s/√n) = (5.8 - 5) / (1.8/√36) = 0.8 / (1.8/6) = 0.8 / 0.3 ≈ 2.67. With df = 35, this would be compared to the t-distribution. The confidence interval or hypothesis test for a mean uses the sample statistic and standard error from the setup.
Why the other 3 choices miss
- A
- The numerator (5 - 5.8) has the wrong sign (should be x̄ - μ₀ = 5.8 - 5 = 0.8). Also, dividing by 1.8/36 = 0.05 instead of 1.8/√36 = 0.3 is an error — the SE uses √n, not n.
- C
- The formula is correct and gives the same numerical answer as A, but the statistic should be called 't', not 'z', because the population standard deviation is unknown and the sample SD (s) is used.
- D
- This divides by s (1.8) alone instead of the standard error s/√n = 1.8/6 = 0.3. The standard error accounts for sample size.
Which of the following is the correct interpretation of a p-value of 0.03 in a one-sample t-test testing H0: μ = 50 vs. Ha: μ ≠ 50?
- AIf the true mean were 50, there is a 3% probability of observing a sample mean as extreme as (or more extreme than) the one observed.Correct
- BThe sample mean differs from 50 by 3 units.
- CThere is a 97% probability that the true mean is not 50.
- DThe probability that the mean is actually 50 is 3%.
Why A is right
The p-value is the probability of obtaining results as extreme or more extreme than the observed data, ASSUMING the null hypothesis is true. A p-value of 0.03 means: if μ really were 50, there is only a 3% chance of seeing a sample mean this far from 50 by random variation alone.
Why the other 3 choices miss
- B
- The p-value is a probability, not a measure of distance in the original units. The difference between the sample mean and μ₀ is measured by the test statistic.
- C
- 1 - p-value is not the probability that H0 is false. This is a common but incorrect interpretation. The p-value does not give the probability of H0 being false.
- D
- P-values do not give the probability that H0 is true. In frequentist statistics, the null hypothesis is either true or false; the p-value describes data probability under H0.
An environmentalist measures mercury levels (ppm) in fish from a lake. A random sample of 20 fish yields: mean = 0.85 ppm, SD = 0.30 ppm. The EPA action level is 1.0 ppm. She tests H0: mu = 1.0 vs. Ha: mu < 1.0 to determine if the mean mercury level is below the action level. The test statistic is t = (0.85 - 1.0)/(0.30/sqrt(20)) = -2.24. With df = 19, the one-sided p-value ≈ 0.019. At alpha = 0.05, what can she conclude?
- AThe test is invalid because mercury levels cannot be normally distributed
- BThere is not sufficient evidence that the mean mercury level is below 1.0 ppm
- CThe mercury level in all fish from this lake is below 1.0 ppm
- DThere is sufficient evidence that the mean mercury level in fish from this lake is below the EPA action level of 1.0 ppmCorrect
Why D is right
The p-value (0.019) is less than alpha (0.05), so we reject H0. There is sufficient evidence to conclude that the mean mercury level in fish from this lake is below the EPA action level of 1.0 ppm. This does not mean every fish is below 1.0 ppm — it refers to the population mean.
Why the other 3 choices miss
- A
- Incorrect because mercury levels can be approximately normal (or the CLT can apply with n = 20 if the data are not severely skewed).
- B
- Incorrect because the p-value (0.019) is less than alpha (0.05), providing sufficient evidence to reject H0.
- C
- Incorrect because the test is about the population mean, not individual fish; some individual fish may still exceed 1.0 ppm.