EPPP Test Research Methods and Statistics Questions and Answers — Questions and Answers
Question 1: A researcher wants to evaluate a new workshop for reducing workplace stress. They randomly assign employees to one of four groups: (1) pretest, workshop, posttest; (2) pretest, no workshop, posttest; (3) no pretest, workshop, posttest; (4) no pretest, no workshop, posttest. This design is best known as which of the following?
- A pretest-posttest control group design
- A Solomon four-group design (Correct answer)
- A factorial design
- A repeated measures design
Correct answer: A Solomon four-group design
The Solomon four-group design is an experimental design that explicitly controls for the potential confounding effect of pretesting. By including two groups that receive the pretest and two that do not, researchers can isolate the impact of the intervention itself from the impact of being tested beforehand.
Question 2: A research team is conducting a study with a small sample size to test a new therapeutic intervention. They are concerned about failing to detect a real effect of the therapy if one truly exists. To minimize this risk, which of the following actions would be most effective?
- Decreasing the alpha level from .05 to .01
- Ensuring the measurement tool for the outcome has low reliability
- Increasing the statistical power of the study (Correct answer)
- Using a one-tailed test instead of a two-tailed test if the direction of the effect is uncertain
Correct answer: Increasing the statistical power of the study
The scenario describes a concern about committing a Type II error (failing to detect a real effect). Statistical power is the probability of correctly rejecting a false null hypothesis, meaning it is the probability of detecting an effect when one exists. Increasing power directly addresses the researchers' concern by reducing the risk of a Type II error. While increasing sample size is a primary way to increase power, ensuring high power is the overarching goal.
Question 3: A psychologist is investigating the relationship between different types of professional training (e.g., CBT, psychodynamic, family systems) and a therapist's primary work setting (e.g., private practice, hospital, community mental health). Which statistical test would be most appropriate to analyze the association between these two variables?
- An independent samples t-test
- A Pearson correlation
- A one-way ANOVA
- A chi-square test of independence (Correct answer)
Correct answer: A chi-square test of independence
Both variables in this scenario, type of professional training and primary work setting, are categorical (nominal) variables. The chi-square test of independence is the appropriate statistical method for examining the relationship or association between two categorical variables.
Question 4: A clinical trial is conducted for a new medication intended to treat social anxiety disorder. The researchers conclude that the medication is not more effective than a placebo. However, in reality, the medication does produce a small but genuine therapeutic effect. The researchers' conclusion is an example of a:
- Type II error (Correct answer)
- Type I error
- Standard error of the mean
- Threat to external validity
Correct answer: Type II error
A Type II error, or a false negative, occurs when a researcher fails to reject a null hypothesis that is actually false. In this case, the null hypothesis (that there is no difference between the drug and placebo) was not rejected, even though a real effect existed.
Question 5: A researcher collects the following data from participants in a weight loss study: (1) their gender, (2) their finishing rank in a 'most improved' competition, and (3) the exact number of pounds they lost. These three variables correspond, respectively, to which scales of measurement?
- Ordinal, Ratio, Nominal
- Interval, Ordinal, Ratio
- Nominal, Ordinal, Ratio (Correct answer)
- Nominal, Interval, Ordinal
Correct answer: Nominal, Ordinal, Ratio
Gender is a categorical variable with no inherent order, which is a nominal scale. Finishing rank has a clear order but the intervals between ranks are not necessarily equal, which is an ordinal scale. The number of pounds lost is a continuous variable with a true zero point (losing zero pounds is meaningful), making it a ratio scale.
Question 6: A research psychologist is planning a study and wants to increase the likelihood of finding a statistically significant result, assuming a true effect exists. All else being equal, which of the following strategies would increase the statistical power of the study?
- Decreasing the effect size
- Decreasing the alpha level (e.g., from .05 to .01)
- Increasing the sample size (Correct answer)
- Using a measure with lower reliability
Correct answer: Increasing the sample size
Statistical power is primarily influenced by three factors: sample size, effect size, and the alpha level. Increasing the sample size is one of the most direct ways to increase statistical power. A larger sample provides a more reliable estimate of the population parameter, making it easier to detect a true effect.
