Deck 9: Evaluating Experimental Procedures

Question 1

The following conditions are important for a study to be considered an experiment. Which one of the following is NOT necessary for an experiment?

A) Treatment is given to an individual or a group.
B) Effects of the treatment are examined.
C) Researchers are examining several groups or individuals.
D) Researchers are trying to establish whether the treatment caused the effects.

Answer 1

C

Question 2

Which one of the following statements about random assignment is INCORRECT?

A) Random assignment to groups ensures that the groups are approximately equal.
B) If the researchers used random assignment, they may not mention it in the report.
C) Random assignment to groups is a key feature of a true experiment.
D) True experiment is the same as a randomized controlled trial.

Answer 2

B

Question 3

A researcher is interested in whether chewing gum during tests helps students to perform better on those tests. Which one of the following scenarios is the best way to do it?

A) Using two existing classes, randomly choose which class will chew gum during tests.
B) Using the same group of students, randomly assign half of the school weeks to include gum chewing on all tests during those weeks.
C) Using 70 students, randomly assign them to two classes, with one of these two classes to use chewing gum during tests.
D) Using twenty existing classes at several local schools, randomly assign half of the classes to chew gum during tests.

Answer 3

C

Question 4

To make statistical comparisons between the groups in an experiment meaningful and to be able to generalize the results, what is the rule of thumb about group size?

A) At least 15 individuals per group.
B) At least 20 individuals per group.
C) At least 30 individuals per group.
D) Group size does not make a difference.

Question 5

What is best way to ensure similarity between groups in an experiment?

A) Random assignment.
B) Pretest.
C) Questionnaire.
D) Gut feeling.

Question 6

For administering a pretest to the groups in a quasi-experiment, all of the following reasons are valid, EXCEPT:

A) Pretest would establish if the groups are similar enough.
B) Pretest would establish baseline scores.
C) Pretest would be used for comparison with post-test to see if the treatment worked.
D) Pretest would help the participants know which topics they need to study for post-test.

Question 7

Regarding experiments that use two pre-existing groups, which one of the following statements is INCORRECT?

A) If the pre-existing groups were not formed at random, the study is a quasi-experiment.
B) If a pre-existing single group is used, with the treatments alternated on different weeks at random, this is a true experiment.
C) For pre-existing groups, a pre-test is needed to compare their similarity.
D) If two pre-existing groups in a quasi-experiment are highly dissimilar initially, it would be hard to interpret the results of the experiment.

Question 8

Which one of the following scenarios for group assignment represents a true experiment?

A) Two pre-existing classes randomly assigned to two different teachers.
B) Three pre-existing classes randomly assigned to three different treatments.
C) Sixty patients assigned to the same treatment on randomly alternating weeks.
D) Sixty police precincts assigned at random to two different policing strategies.

Question 9

What is the biggest problem with a single-group experimental design that can affect the validity of the results?

A) The experimenters are not sure how to alternate the treatments.
B) The participants are not sure why things keep changing back and forth.
C) The participants are not sure if they are in a study or not.
D) The participants are not sure if they are in an experimental or control group.

Question 10

Regarding the treatments administered in an experiment, which one of the following statements is INCORRECT?

A) In a research report, the treatments should be described in enough detail.
B) Experiments should only use the treatments previously validated in other published research.
C) The treatments should be administered by trained personnel.
D) The administration of treatments should be monitored through observation or questioning.

Question 11

What is the experimental condition that we must worry about the most in terms of the personal effect?

A) If each treatment is administered by a different person.
B) If each treatment is administered by the same person.
C) If the person administering the treatments is also the person responsible for randomly assigning participants to the treatments.
D) If the person testing the results also knows who is in the experimental group and who is in the control group.

Question 12

A researcher is planning an experiment to test whether a new teaching approach is helpful in improving the results of peer tutoring. Which one of the following procedures is LEAST likely to reduce the effects of confounding variables?

A) Randomly assign half of the peer tutors to implement the new teaching approach.
B) Randomly assign research assistants to assess the results of the new teaching approach.
C) Make sure the students being tutored are not aware of whether their peer tutor is using the new approach.
D) Make sure the research assistants assessing the results of the experiment don't know which student received which teaching approach.

Question 13

What is a double-blind experiment?

A) Neither experimental nor control group participants are aware of their group assignment.
B) Neither the researcher nor the participants are aware of the group assignment.
C) Neither the person assessing the results nor the participants are aware of the group assignment.
D) The participants do not know that they are participating in an experiment and the researcher does not know which type of experiment it is.

Question 14

All of the following scenarios have been implemented as true experiments. Which one is LEAST likely to present the risks of demand characteristics?

A) Testing the effects of team-teaching versus a usual single-instructor teaching approach on students' test scores.
B) Testing the effects of a court-mandated substance abuse treatment versus probation on repeat offending.
C) Testing the effects of playing violent versus non-violent video games on geometry test scores.
D) Testing the effects of a video on drinking-and-driving dangers on the attitudes about driving under the influence.

Question 15

Natural experiments are most likely to present problems with:

A) external validity.
B) internal validity.
C) demand characteristics.
D) attrition.

Question 16

A researcher conducted a true experiment testing whether a low-sugar diet versus a low-fat diet leads to a more substantial weight loss. If this experiment has high internal validity, what does it mean?

A) The participants did not know whether they were assigned to the low-sugar or low-fat diet.
B) The experiment was conducted in natural settings.
C) The experiment has established the effect of the diet type on weight-loss outcomes.
D) The results of the experiment can be generalized to the entire population.

Question 17

Which one of the following conditions helps achieve high internal validity in an experiment?

A) Random assignment to experimental and control groups.
B) Random selection of participants for an experiment.
C) Experiment conducted in natural settings.
D) Experiment where participants are free to withdraw at any time.

Question 18

All of the following factors increase the risk of attrition, EXCEPT:

A) longitudinal studies.
B) experiments with complex tasks.
C) experiments of long duration.
D) experiments with volunteer participants.

Question 19

All of the following statements indicate the risks that attrition presents for an experiment, EXCEPT:

A) People dropping out of the experimental group may have different characteristics than people dropping out of the control group.
B) Participants in any experiment can withdraw at any time so attrition risks are the same for all experiments.
C) Attrition can especially bias the results of an experiment that places high demands on its participants.
D) Differential attrition can affect the generalizability of the results of a long experiment.