<!DOCTYPE html>

Chapter 8 - Practice Problems

Implement necessary functions

Problem set

Data for Problems 8.2 - 8.13

8.2 Test for a significant difference between the variances.

H0: var1 == var2
Ha: var1 != var 2

The question is, what's the F statistic of the F distribution with those parameters (alpha/2 and 1- alpha/2)?

We want a function where we can input the probability (alpha) and get out the corresponding F-statistic... That means we want the PPF.

So our test statistic falls within the acceptance region of the F distribution. This means that, given that the null hypothesis is true, it is highly likely to get these results, therefore we cannot reject (i.e., we can accept) the null hypothesis that the two variances are equal.

8.3 What is the appropriate procedure to test for a signifi- cant difference in means between the two groups?

Since we discovered that the two samples come from populations with equal variances, the Two samples independent t-test with equal variances is the most appropriate test here.

8.4 Implement the procedure in Problem 8.3 using the critical-value method.
8.5 What is the p-value corresponding to your answer to Problem 8.4?

So the p value is 0.19, which means we don't have a significant difference in means between the two samples, and that they do not represent two distinct populations with distinct means (given a 5% signficance threshold).

8.6 Compute a 95% CI for the difference in means between the two groups.

*8.7 Suppose an equal number of 12- to 14-year-old girls below and above the poverty level are recruited to study differences in calcium intake. How many girls should be recruited to have an 80% chance of detecting a significant difference using a two-sided test with α = .05?

Equation for computing the Sample Size Needed for Comparing the Means of Two Normally Distributed Samples of Equal Size Using a Two-Sided Test with Significance Level α and Power 1 − β:

INSERT EQUATION HERE

so ~134 people should be recruited for this study to have an 80% chance of detecting an effect (assuming that there is one). We already know there is no significant difference between the means though.

*8.8 Answer Problem 8.7 if a one-sided rather than a two- sided test is used.

Intuitively it makes sense that we require less people since a two-tailed test is harder to pass, so it would require a higher number of people to achieve the same statistical power.

*8.9 Using a two-sided test with α = .05, answer Problem 8.7, anticipating that two girls above the poverty level will be recruited for every one girl below the poverty level who is recruited.

Sample Size Needed for Comparing the Means of Two Normally Distributed Samples of Unequal Size Using a Two-Sided Test with Significance Level α and Power 1 − β:

INSERT EQUATION HERE

*8.10 Suppose 50 girls above the poverty level and 50 girls below the poverty level are recruited for the study. How much power will the study have of finding a significant difference using a two-sided test with α = .05, assuming that the population parameters are the same as the sample estimates in Problem 8.2?

Almost no chance at all of detecting a significant difference, which makes sense, I think, given the difference of the means of the samples (quite small) and the sample standard deviations (large, relative to the difference in means).

*8.11 Answer Problem 8.10 assuming a one-sided rather than a two-sided test is used.

*8.12 Suppose 50 girls above the poverty level and 25 girls below the poverty level are recruited for the study. How much power will the study have if a two-sided test is used with α = .05?

*8.13 Answer Problem 8.12 assuming a one-sided test is used with α = .05.