Question 1 : Which of the following correctly describes the difference between paired and independent sample data?
a. Independent samples operate based on a known σ, whereas paired samples may operate with a sample standard deviation.
b. Independent samples are used to address the difference between two separate populations, whereas paired samples involve a matched analysis.
c. Independent samples incorporate a t distribution, whereas paired samples may use the z distribution for approximation.
d. Independent samples always have the same sample size, whereas paired samples may compare groups with different sample sizes.
Question 2 : Which of the following observes a comparison between two independent samples?
a. The comparison between performance on a board certification exam prior to, and post implementation of a new preparation program among a group of surgeons.
b. Observing the difference in fasting glucose levels between a group of individuals receiving a new study drug and a group receiving a placebo.
c. Determination if a SRS of children in a classroom is performing up to a particular level on a new standardized exam.
Question 3 : Which of the following does not observe a comparison between two independent samples?
a. Observation of the difference in forced vital capacity between a group of children living in a home with above a specific level of mold and a group of children living in a home with no mold.
b. Determining if a group of patients receiving a novel vaccine for Hepatitis C are treated versus those receiving best standard of care.
c. Determining the difference in performance on an academic exam in a SRS of children matched on a variety of various characteristics.
Question 4 : Which of the following is not required for two-sample procedures?
a. Independence within samples
b. Independence between groups
c. Both samples should be taken from an SRS or equivalent
d. Normality for at least one of the samples
Question 5 : When is it not appropriate to assume equal variances in a population?
a. When the populations' standard deviations are slightly different
b. When the results following the hypothesis test with a null hypothesis σ1=σ2 conclude in a pvalue less than 0.05
c. It is almost always appropriate to assume equal variances
d. Whenever the variances are different by more than 10%
Question 6 : Suppose a researcher is interested in the effectiveness of a new therapy at reducing scores on the global depression scale (GDS). Assume that an effective change in the GDS scale between placebos and treatment groups must be more than 5 points. Which of the following correctly identifies the null and alternative hypotheses for the test for independent means?
a. Ho: μ1 - μ2 = 5, Ha: μ1 - μ2 ≠ 5
b. Ho: μ1 - μ2 > 5, Ha: μ1 - μ2 = 5
c. Ho: μ1 - μ2 < 5, Ha: μ1 - μ2 = 5
d. Ho: μ1 - μ2 = 5, Ha: μ1 - μ2 > 5
Question 7 : The company that sponsors the MCAT examination for hopeful MDs will be changing the examination format in order to remove one section and add in a section with biochemistry content in the next few years. The sponsors have decided to perform a pilot analysis of the new format in order to see if there is a significant difference in overall scoring. Assume that the total score does not change. If the researchers find a mean difference above 15, then the new content may have to be altered due to an excessive difficulty increase. Which of the following represents the correct set of null and alternative hypotheses?
a. Ho: μ1 - μ2 >15, Ha: μ1 - μ2 ≠ 5
b. Ho: μ1 - μ2 = 15, Ha: μ1 - μ2 >15
c. Ho: μ1 - μ2 < 15, Ha: μ1 - μ2 = 15
d. Ho: μ1 - μ2 = 15, Ha: μ1 - μ2 < 15
Question 8 :Which of the following accurately describes the relationship between the various parameters incorporated in the formula for calculating the power of an independent sample t-test?
a. As the sample size increases, the power of your test decreases
b. As the null value for the difference in averages between groups decreases, so too does the power
c. As the Z value approaches virtual certainty, the power of your test will always increase
d. As the pooled standard deviation estimate increases, your power decreases
Question 9 : Which of the following is an incorrect pairing of a parametric test to its nonparametric equivalent?
a. One Way Analysis of Variance: Kruskal-Wallis Test
b. Independent t-test: Mann-Whitney rank Sum Test
c. One sample t-test: Spearman's rank correlation coefficient
d. One sample t-test: Wilcoxon signed rank test
Question 10 : Which of the following describes the procedure for calculating a test statistic in the Kruskal-Wallis procedure?
a. Observing the difference in medians between the distributions
b. Ranking the observations and determining if there is a significant difference in mean rank scores
c. Calculating the mean difference between the groups
d. Using a pooled variance to determine if the variances between groups significantly differ
Question 11 : Which of the following accurately describes the family-wise error rate?
a. As the number of comparisons being made increases, the type I error rate decreases
b. As the number of comparisons being made decreases, the type I error rate increases
c. As the number of comparisons being made increases, so does the type II error rate
d. As the number of comparisons being made increases, so does the type I error rate
Question 12 :Which of the following correctly describes the null and alternative hypotheses for an analysis of variance between three treatment groups (µ1 µ2 µ3)?
a. Ho: µ1= µ2, Ha: µ1≠ µ2
b. Ho: µ1 = µ2 = µ3, Ha: µ1≠ µ2
c. Ho: µ1 = µ2 = µ3, Ha: At least 2 means differ
d. Ho: µ1 = µ2 = µ3, Ha: All means differ
Question 13 : Which of the following is not an assumption that is required for the analysis of variance procedure?
a. Sampling independence among groups
b. Homoscedasticity in populations
c. Heteroscedasticity in populations
d. Normality of sampling distributions of means
Question 14 : Which of the following procedures requires an F statistic in order to determine a pvalue?
a. One way analysis of variances
b. Two sample independent t-test
c. Lease squares difference
d. Equality of variance
e. Both A and D
Question 15 : Which of the following is true of the F statistic calculated in the one way analysis of variance?
a. The F statistic is calculated by dividing the mean square within by the mean square between variances
b. The F statistic involves a singular degrees of freedom calculation approximated by the Welch equation
c. Large F statistics suggest the expected mean differences are not merely due to chance
d. Large F statistics suggest the observed mean differences are not merely due to chance