Free EPPP Research Methods and Statistics Practice Questions

The Research Methods and Statistics content area represents 7% of the EPPP (Part 1-Knowledge) and covers sampling, research designs, statistical analysis and interpretation, validity threats, and program evaluation. Review each answer and explanation as you practice.

The EPPP format is 225 questions in 255 minutes; this page is a focused content-area sample.

What Research Methods and Statistics Covers on the EPPP

ASPPB's test specifications for this content area include:

How to Study Research Methods and Statistics

EPPP Research Methods and Statistics Sample Questions with Answers

Sample Question 1 — Research Methods and Statistics

A single group of adults completes a measure of attitudes toward mental health treatment, attends a 10-week stigma-reduction program, and completes the measure again. During those 10 weeks a celebrity's widely covered disclosure of depression dominates the news. Which threat to internal validity is MOST relevant?

  1. A. Testing
  2. B. Statistical regression
  3. C. Maturation
  4. D. History (Correct answer)

Correct answer: D

Explanation: D is correct because an outside event that occurs between pretest and posttest and could itself change attitudes is a history threat, and a one-group design cannot separate it from the program's effect. A is incorrect because testing refers to changes caused by taking the pretest itself, not by events in the wider environment. B is incorrect because regression applies when participants are selected for extreme scores, and this group was not selected that way. C is incorrect because maturation refers to natural changes within participants over time, such as growing older or more tired, rather than an outside event.

Sample Question 2 — Research Methods and Statistics

A researcher changes the significance level for a study from .01 to .05 while keeping the sample size the same. What is the effect on the study?

  1. A. Power increases, and the probability of a Type I error increases (Correct answer)
  2. B. Power is unchanged, because only sample size affects power
  3. C. Power decreases, and the probability of a Type I error increases
  4. D. Power increases, and the probability of a Type I error decreases

Correct answer: A

Explanation: A is correct because a less strict alpha makes it easier to reject the null hypothesis, which increases power (and lowers Type II error risk) while raising the chance of rejecting a true null hypothesis. B is incorrect because power depends on alpha and effect size as well as sample size. C is incorrect because a less strict alpha makes rejecting the null easier, which increases rather than decreases power. D is incorrect because a less strict alpha raises, not lowers, the probability of rejecting a true null hypothesis.

Sample Question 3 — Research Methods and Statistics

A researcher divides a population of clinic patients into ethnic groups and then randomly samples from each group in proportion to that group's share of the population. This procedure is called:

  1. A. Single-stage cluster sampling
  2. B. Stratified random sampling (Correct answer)
  3. C. Quota sampling by subgroup
  4. D. Systematic random sampling

Correct answer: B

Explanation: B is correct because the population is divided into subgroups (strata) and random samples are drawn from each in proportion to its size, which guarantees each subgroup is represented accurately. A is incorrect because cluster sampling randomly selects whole naturally occurring groups, such as clinics or schools, rather than sampling individuals within every subgroup. C is incorrect because quota sampling fills subgroup quotas with whoever is available, so selection within each subgroup is not random. D is incorrect because systematic sampling selects every kth person from a list and does not first divide the population into subgroups.

Sample Question 4 — Research Methods and Statistics

To study adults who inject drugs and are not in treatment, a researcher interviews a few people known to an outreach worker and then asks each participant to refer others they know who meet the study criteria. Which sampling method is being used?

  1. A. Cluster sampling
  2. B. Convenience sampling
  3. C. Systematic random sampling
  4. D. Snowball sampling (Correct answer)

Correct answer: D

Explanation: D is correct because recruiting through referrals from existing participants is snowball sampling, which is useful for hidden or hard-to-reach populations that have no sampling frame. A is incorrect because cluster sampling randomly selects whole natural groups from a list of such groups, which requires a sampling frame this population lacks. B is incorrect because convenience sampling recruits whoever is readily available, such as clinic attendees, without relying on participants to refer others. C is incorrect because systematic sampling picks every kth name from a complete list of the population, and no such list exists here.

Sample Question 5 — Research Methods and Statistics

A researcher randomly selects 400 adults from a national registry. Each participant then chooses whether to join a mindfulness group or a waiting list, and the mindfulness group later reports less stress. Which conclusion is BEST supported?

  1. A. Because both selection and assignment were random, the study has strong internal and external validity
  2. B. Results may generalize to the registry population, but self-selection could explain the group difference (Correct answer)
  3. C. Because participants were selected at random, the group difference can be attributed to the mindfulness training
  4. D. Because assignment was not random, the results cannot be generalized to the registry population at all

Correct answer: B

Explanation: B is correct because random selection supports external validity (generalizing to the registry population), but without random assignment the groups may differ in motivation or other traits, which threatens internal validity. A is incorrect because participants chose their own condition, so assignment to groups was not random. C is incorrect because random selection affects generalizability, while causal inference about the training requires random assignment to conditions. D is incorrect because generalizability depends on how the sample was selected, and random selection from the registry supports it.

Sample Question 6 — Research Methods and Statistics

Which of the following variables is measured on a ratio scale?

  1. A. Reaction time in milliseconds (Correct answer)
  2. B. Order of finish in a race
  3. C. A client's primary diagnosis
  4. D. Temperature in degrees Fahrenheit

Correct answer: A

Explanation: A is correct because reaction time has equal intervals and a true zero point, so ratios are meaningful (400 ms is twice as long as 200 ms), which defines a ratio scale. B is incorrect because finishing order ranks runners without equal distances between ranks, which makes it an ordinal scale. C is incorrect because diagnostic categories are labels with no inherent order, which makes them a nominal scale. D is incorrect because Fahrenheit temperature has equal intervals but an arbitrary zero, which makes it an interval scale.

Sample Question 7 — Research Methods and Statistics

A state agency wants to survey high school students about vaping. It randomly selects 30 of the state's 400 high schools and surveys every student in each selected school. Which sampling method is this?

  1. A. Cluster sampling (Correct answer)
  2. B. Stratified random sampling
  3. C. Simple random sampling
  4. D. Convenience sampling

Correct answer: A

Explanation: A is correct because the unit randomly selected is a naturally occurring group (a school), and all members of each selected group are included, which is cluster sampling. B is incorrect because stratified sampling draws individuals at random from every subgroup, whereas here most schools are not sampled at all. C is incorrect because simple random sampling gives every individual student an equal and independent chance of selection, rather than selecting whole schools. D is incorrect because the schools were chosen at random, not because they were easy to reach.

Sample Question 8 — Research Methods and Statistics

Clinicians rate each client's symptom severity as none, mild, moderate, or severe. Using all the information these ratings contain, which measure of central tendency is MOST appropriate?

  1. A. The median (Correct answer)
  2. B. The mean
  3. C. The mode
  4. D. The trimmed mean

Correct answer: A

Explanation: A is correct because the ratings are ordinal, meaning they are ordered but the distances between categories are not known to be equal, and the median uses that order without assuming equal intervals. B is incorrect because the mean assumes equal intervals between scale points, which ordinal ratings do not provide. C is incorrect because the mode is permissible but uses only category frequencies and ignores the order information that ordinal data contain. D is incorrect because a trimmed mean is still an average of scores and, like the mean, assumes interval-level measurement.

Sample Question 9 — Research Methods and Statistics

In an ABAB single-subject design, what is the main purpose of the second A phase?

  1. A. To raise the performance criterion in gradual steps
  2. B. To show that behavior reverts when treatment is withdrawn (Correct answer)
  3. C. To establish a stable level of behavior before any treatment
  4. D. To measure whether the behavior generalizes to new settings

Correct answer: B

Explanation: B is correct because withdrawing treatment and seeing the behavior revert, then improve again when treatment is reinstated, shows that the treatment rather than chance events controls the behavior. A is incorrect because stepwise changes in the criterion define the changing criterion design. C is incorrect because that is the purpose of the first A phase, the initial baseline. D is incorrect because generalization is assessed by measuring the behavior in other settings, not by withdrawing treatment.

Sample Question 10 — Research Methods and Statistics

A psychologist wants to test a program that teaches sight-word reading to three children. Once the words are learned, reading them is not expected to disappear if teaching stops. Which design is MOST appropriate?

  1. A. A changing criterion design with each child
  2. B. A simple AB design with one child
  3. C. A multiple baseline across the three children (Correct answer)
  4. D. An ABAB reversal design for each child

Correct answer: C

Explanation: C is correct because because the skill will not reverse, a withdrawal design cannot work; starting teaching at different times for each child and showing that each improves only when teaching begins demonstrates control. A is incorrect because a changing criterion design shifts a performance goal in steps for one behavior, and it does not use staggered starts to rule out outside events. B is incorrect because an AB design has no replication, so outside events or maturation could explain any improvement. D is incorrect because a reversal design depends on the behavior returning to baseline when treatment is withdrawn, which a learned reading skill will not do.

Keep Practicing

Take the 10-question EPPP quick-start practice exam across all eight content areas, or return to the EPPP practice exam hub.