Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
37 Cards in this Set
- Front
- Back
Maximize the power of the correlation coefficient by__
|
minimizing error variance, avoiding restricted range, testing a large N, and computing parametric, linear statistics when appropriate.
|
|
Three ways a correlation coefficient is used to sow reliability
|
1. inter-rater reliability
2. test-retest reliability 3. split-half reliability |
|
inter-rater reliability
|
To demonstrate that the behavior is judged reliably, which means that we have highly consistent ratings.
|
|
test-retest reliability
|
It indicates that participants tend to obtain the same score when tested at different times.
|
|
split-half reliability
|
It indicates that participants' scores on some trials consistently match their scores on other trials. Ex. split a test in half by comparing the odd-numbered trials to the even-numbered trials.
|
|
Test-retest reliability is the correlation between repeated testings, and split-half is the correlation between different trials within one test.
|
A coefficient of +.80 or higher is usually required for the procedure to be considered reliable.
|
|
face validity
|
is the extent to which a measurement procedure appears to measure what it's intended to measure.
|
|
Convergent validity
|
is the extent to which the scores obtained from one procedure are positively correlated with scores obtained from another procedure that is already accepted as valid.
|
|
Discriminant validity
|
Is the extent to which the scores obtained from one procedure are not correlated with scores from another procedure that measure other variables or constructs.
|
|
With convergent validity, a procedure correlates with another procedure that is valid. With discriminant validity, a procedure does not correlate with other, unintended measures.
|
n/a
|
|
Criterion validity
|
Is the extent to which a procedure can distinguish between participants on the basis of some behavior.
|
|
Two types of criterion validity
|
1. concurrent validity
2. predictive validity |
|
Concurrent validity
|
is the extent to which a procedure correlates with an individual's present behavior.
|
|
predictive validity
|
is the extent to which a procedure correlates with an individual's future behavior.
|
|
criterion validity is the extent to which a procedure relates to a specific behavior, either distinguishing the behavior concurrently or predicting future behavior.
|
n/a
|
|
closed-ended question
|
the researcher provides the alternatives from which the participant selects. (may include, multiple-choice, true-false, yes-no, and rating-scale questions)
|
|
open-ended question
|
the participant determines both the alternatives to choose from and the response. (written question, essay question)
|
|
projective test
|
the participant describes or interprets and ambiguous stimulus.
|
|
Rorschach test
|
in which participants interpret ambiguous inkblots
|
|
Thematic Apperception test
|
in which they interpret ambiguous picture
|
|
content analysis
|
we score a participant's written or spoken answer by counting specified types of responses.
|
|
structured interview
|
participants are asked specific, predetermined questions in a controlled manner.
|
|
unstructured interview
|
in which the researcher has a general idea of the open-ended questions that will be asked, but there is freedom of discussion and interaction between participant and interviewer.
|
|
interviews are preferred when the researcher must react to a participant's responses, but questionnaires are more reliable and less susceptible to demand characteristics.
|
The goal of question construction is to reliably and validly discriminate between participants on the variable being studied.
|
|
Avoiding three types of questions
|
1. double-barreled question
2. leading questions 3. Barnum statements |
|
double-barreled question
|
these are questions that have more than one component
|
|
leading questions
|
these are questions that communicate social desirability or experiment expectancies so that there is only one obvious response.
|
|
Barnum statements
|
These are questions that are so global and vague that everyone would agree with them or select the same response for them.
|
|
A question should be a clear, precise, and unbiased statement of a single idea, to which different participants are likely to respond differently
|
n/a
|
|
response scale
|
which is the number and type of choices to provide for each question.
|
|
likert-type questions
|
when measuring responses that fall along a continuum. These consist of a declarative statement accompanied by a rating scale.
|
|
Techniques for dealing with order effects
|
1. provide practice questions
2. use funnel questions 3. use filter questions 4. counterbalance order effects 5. prevent response sets 6. use alternate forms |
|
funnel questions
|
are general questions that lead to more specific questions. Order questions from general to specific.
|
|
filter questions
|
are general questions that determine whether participants should answer additional, detailed questions.
|
|
counterbalance order effects
|
balance the effects of one particular order by creating different orders of questions for different participants so that questions that appear early in some questionnaires appear later in others and vice versa
|
|
Prevent response sets
|
To prevent rote responding, vary the question format to try to force participants to read and think about each question.
|
|
alternate forms
|
are different versions of the same questionnaire give to different participants.
|