Writing and Evaluating Test Items Flashcards

1
Q

What’s the most common used method in psychological testing?

A

Questionnaires

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Questionnaire characteristics (3)

A

(1) Written series of questions
(2) Structured stimuli (i.e. questions)
(3) Structured responses (i.e. response format)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Questionnaires Advantages (4)

A

(1) Presentation of stimuli is well controlled
(2) Scoring highly reliable
(3) Efficient to administer to large numbers
(4) Inexpensive

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Example of Questionnaires with Dichotomous Choice Formats

A

True-False and Agree-Disagree

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Dichotomous formats are often seen in ______ tests

A

Personality, e.g. MMPI

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Compared to other formats, the dichotomours format is ________

A

less reliable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Rating Scales Formats def

A

Category Format. Ask responder to evaluate/rate something along a defined continuum.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What’s the problem with rating scales formats?

A

Number of points? How many options?
More options = more variability
At what point is it too much?
Middle point? Acknowledges that pple might NOT have an opinion, but can be an easy way out
Often 10, between 4 and 7 it’s good.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

How many points are optimal for a likert scale?

A

If likert scale, 7 = optimal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Forced Choice Formats

A

Person is presented with 2 to 4 stimuli and asked to choose among them.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Q-sort Formats

A

Forced distribution of items into categories
E.g. E.g. Give person list of 100 characteristics. Group the characteristics according to how like the person the characteristics are:

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Four Steps in the Question-Answer Process

A

(1) Comprehension: Attending to questions and instructions
(2) Retrieval: Retrieval of relevant information
(3) Judgment: Integration of retrieved information
(4) Response: Mapping the judgment on the response category

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Issues for Questionnaires (3)

A

RESPONSE SET: Tendency for people to respond to questions in a way that paints a certain picture of themselves instead of providing honest answers
(1) Acquiescence = Tendency to agree, say true, say often.
(2) Social desirability = Tendency to present self in a socially favorable manner
(3) Random responding = Ignoring or paying insufficient attention to item content

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Errors in completing a questionnaire = ____ (def)

A

Formally valid answers that do NOT reflect true scores, undermining data quality

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

How can we combat Acquiescence bias?

A

Use reverse-score items

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What are the 2 components of Social desirability?

A

Impression management & Self deception

17
Q

How can we combat social desirability? (3)

A

(1) Measure influence: assess discriminant validity
-> Discriminant validity evaluates whether a test measures what it’s supposed to (its intended construct) and not something else (like social desirability).
-> If your measure overlaps too much with social desirability, it indicates the test may be influenced by this bias rather than the true construct.
(2) Marlowe-Crowne social desirability scale
(3) Change response format (forced choice; Q-sort)

18
Q

Random responding occurs when respondents ____ (3)

A

(1) Do not read an item
(2) Do not understand an item
(3) Are unmotivated to think about the item

19
Q

How can we detect random responding? (4)

A

(1) INSTRUCTED response items: Ask for a specific answer (e.g. “choose strongly disagree”)
(2) BOGUS items: Ask about impossible or improbable scenarios (e.g. “I was born before 1920”)
(3) SELF-REPORT items: Ask participants about their care and engagement DURING the survey
(4) RESPONSE TIME: Computed after data collection but must be considered before starting

20
Q

Writing Good items (12)

A
  • Single idea per item stem
  • Write each item in a clear and direct manner
  • Avoid long items
  • Avoid double negatives
  • Reading level appropriate for intended test-takers
  • Avoid slang or colloquial language
  • Make all items independent
  • Ask someone else to review items to reduce ambiguity and inaccuracies
  • Make all responses similar in length and detail
  • Make sure the item has only one best answer
  • Avoid words such as “always” and “never”
  • Avoid overlapping responses