Experimental design 2/surveys, questionnaires Flashcards

1
Q

What is the Likert Scale and what makes it useful?

A

he format of a typical five-level Likert item, for example, could be:
Strongly disagree
Disagree
Neither agree nor disagree
Agree
Strongly agree

Its useful because the answers can be quantified and subsequently parametric statistics perfomed (mean, SD etc.)

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

What is the difference between the inductive approach and deductive approach to scale construction? When would you use which?

A

Inductive Approach = generate items of scale first (used when unfamiliar topic)
Deductive Approac = use theoretical definitions of construct to generate items

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

What is a Factor Analysis?

A

a statistical method to describe variability among observed correlated variables by using unobserved variables called factors during scale construction

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

What is an Exploratory factor analysis

A

reduce set of observed variables to a smaller more parsimonious set of variables

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

What is Confirmatory Factor Analysis?

A

asses the factor structure for overall validity

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

During Scale Exploratory Factor Analysis when are items removed?

A

When they do not load onto any factors

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

What characterizes a good psychrometric scale (in statistical terms)?

A

items on one scale should be highly correlated and less corelated with other scales, good items have a HIGH variance (regarding answers to thaat one item)

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

What is used to asses the internal consistency of a scale?

A

Cronbach’s alpha = degree of interrelation of items of a scale.

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

Does there always have to be a replication at the end of scale construction in order to validate the scale?

A

Well if the question is asked like this: obv. not.

Replication at end with different sample
OR split group in half
OR use confirmatory factor analysis for replication

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