WEEK 4: Foundations of Research Flashcards
What refers to the consistency or stability of your measurements?
A. Reliability
B. Validity
Reliability
Results are consistent, but do not measure the intended variable
A. High reliability, high validity
B. High reliability, low validity
C. Low reliability, low validity
D. Low reliability, High validity
High reliability, low validity
Results measure the intended variable, and measure them very consistently.
A. High reliability, high validity
B. High reliability, low validity
C. Low reliability, low validity
D. Low reliability, High validity
High reliability, high validity
Results that have low reliability (i.e. vary widely) are therefore also low in validity.
A. High reliability, high validity
B. High reliability, low validity
C. Low reliability, low validity
D. Low reliability, High validity
Low reliability, low validity
Results that have low reliability (i.e. vary widely) are therefore also low in validity.
A. High reliability, high validity
B. High reliability, low validity
C. Low reliability, low validity
D. Low reliability, High validity
Low reliability, low validity
This combination is not actually possible —it does not exist.
A. High reliability, high validity
B. Low reliability, High validity
C. Low reliability, low validity
D. Low reliability, High validity
Low reliability, High validity (Test yourself: why is it impossible to have high validity if you have low reliability?)
What occurs when not all participants who begin a study complete the study?
A. Competence effect
B. History
C. Regression to the mean
D. Attrition
Attrition
Attrition is more common in longitudinal studies
What is a general term for an event that occurs after the beginning of the study, and before the post-test measurements are taken.
A. Competence effect
B. History
C. Regression to the mean
D. Attrition
History
(For this example, you are running a trial of a new psychological intervention in prisons. Your participants are taking part in a new psychoeducational class, one which aims to decrease their overall distress levels by teaching them how to manage their own distress)
In plain terms, means that extreme results tend not to be repeated. If you have an extreme result, it is highly likely that on repetition, the result will be closer to the average, or mean..
A. Competence effect
B. History
C. Regression to the mean
D. Attrition
Regression to the mean
(As one example, let’s say that you meet a new client who is very highly distressed in your first session together. Based on this tendency—regression toward the mean—it is unlikely that he will be as extremely distressed in your second session. His distress levels have probably fallen closer to the mean, or average, distress level in the population).
What can be either a participant effect—a change in the level of competence of the participant—or an experimenter effect—a change in the level of competence of the experimenter. Both can affect results.
A. Competence effect
B. History
C. Regression to the mean
D. Attrition
Competence effect
(For example, a participant who performs poorly on a test the first time will probably improve upon taking the same test again. (Furthermore, if they performed extremely poorly, regression toward the mean would predict that they would improve significantly on taking the test again).
Selecting your group randomly is known as what?
A. Random sampling
B. Non-random sampling
C. Attrition
D. Mean
Random sampling (Selecting your group randomly is known as 'random sampling')
(For example, if you chose a group of female participants and interviewed them about their feelings regarding childbirth, you would see one kind of trend, but if you asked a random group of male and female participants, you would probably see a different trend.)
Selecting your group according to criteria rather than randomly is known as what?
A. Random sampling
B. Non-random sampling
C. Attrition
D. Mean
Non random sampling
(Because non-random sampling has the potential for bias, you may wonder why non-random sampling techniques are used at all)
Because non-random sampling has the potential for bias, you may wonder why non-random sampling techniques are used at all. It is true that they tend to be weaker samples, and that therefore you often cannot infer as much from the results.
