Chapter 5

Question 1

What are the 3 types of measures?

Answer 1

1. Self-report
2. Physiological
3. Observational

Question 2

What is operationalisation?

Answer 2

The process of turning a construct of interest into a measured or manipulated variable

Question 3

What are the two definitions of variables in physiological research?

Answer 3

The conceptual definition and the operational definition

Question 4

How are variables defined in research?

Answer 4

A conceptual definition is first made of the idea that we wish to measure (such as happiness being subjective well-being) and a operational definition is then made to measure the variable (the quantifiable measure). These definitions can define the same concept in many different ways.

Question 5

How do self-report measures operationalise variables?

Answer 5

It records people’s answers to questions about themselves in a questionnaire or interview

Question 6

How are self-report measures used on children?

Answer 6

Self reports may be replaced by parent reports or teacher reports

Question 7

How do observational measures operationalise variables?

Answer 7

It records observable behaviours or physical traces of behaviours

Question 8

How do physiological measures operationalise variables?

Answer 8

By recording biological data - usually with equipment to amplify, record or analysize the data. These measures may record high IQ as their brains dont light up as much on complex problems

Question 9

How do you know which operationalisation is the best?

Answer 9

Variables are often measured by a multitude of different measures so there is no necessary “best”. A way you can tell if a variable is measured well is if the results that measure an idea in all 3 measures follow the same pattern.

Question 10

What are the different operational variables?

Answer 10

Categorical (although they have categories, they can be represented by numbers but they often dont mean anything) and quantitative (they have meaningful numbers)

Question 11

What are the three types of quantitative variables?

Answer 11

Ordinal, ratio and interval scales

Question 12

What makes ordinal scales different from the others?

Answer 12

Ordinal scales focus on rank and often the only meaningful part of the data is the differences - but not the quantifiable differences.

Question 13

What are the requirements of an interval scale?

Answer 13

1. The numbers need to represent equal intervals between levels
2. There is no “true” zero aka zero does not mean there is nothing there

Question 14

What are the requirements of ratio scales?

Answer 14

When the numbers have equal intervals and when the value of zero truly means there is nothing there.

Question 15

What are the 2 aspects of construct validity?

Answer 15

Reliability (how consistent) and validity (does it measure what its supposed to)

Question 16

What are the three kinds of reliability?

Answer 16

1. Test-retest reliability (same score measured on repeated tests)
2. Interrater reliability (consistent scores no matter who measures)
3. Internal reliability (a consistent pattern of answers given no matter the question)

Question 17

When does test-retest reliability apply?

Answer 17

It applies in self-report, observational or physiological measures - but it is most effective at measuring constructs that are stable such as IQ

Question 18

When does interrater reliability apply?

Answer 18

It is the most relevant for observational measures

Question 19

When does internal reliability apply?

Answer 19

It applies to measures that combine multiple items

Question 20

What are two statistical devices researchers can use?

Answer 20

1. Scatterplots
2. Correlation coefficient r

Question 21

How is evidence for reliability an association claim?

Answer 21

It is between the association at an earlier time and a later time, one coder and another and one version of a measure and another

Question 22

How can scatterplots and r be used to measure reliability?

Answer 22

By comparing the measurements taken at different times, or by different people or otherwise, if they data points flow in a straight line and have a high r, that suggests strong reliabilty (although not perfect cause measurement error)

Question 23

How is test-retest reliability assessed with r or a scatterplot?

Answer 23

2 measures taken at 2 different times and compared on a scatterplot - only applicable for traits that do not really change

Question 24

How is interrator reliability assessed with r or a scatterplot?

Answer 24

2 measures taken by 2 different researchers - a negative correlation would be worse than just a low r. Kappa is used instead of r for categorical variables .

Question 25

How is internal reliability assessed?

Answer 25

Often by the same question being given worded in multiple ways and summed to create a single composite score.

Question 26

How is internal reliabiltiy statistically measured?

Answer 26

Researchers compute correlations between every item and every other item where the Average Inter-item correlation (average of all these correlations), if between .15 and .5 often means they go well together. Researchers my also use Cronbach’s alpha which combines the AIC and number of items, with a strength scale the same to r (strong is 0.8 < x)

Question 27

When is construct validity very important?

Answer 27

When a construct is not directly observable such as happiness or IQ

Question 28

What are three different kinds of validity?

Answer 28

1. Face validity (subjectively considered a plausible operationalisation)
2. Content validity (measure must capture all parts of a defined construct)
3. Criterion validity (the measure is associated with a concrete behavioural outcome that is should be associated with)

Question 29

How can criterion validity be measured?

Answer 29

Through scatterplots and r where the behavioural outcome and the measure should have a high r to have validity. It is good for self-report measures as it indicates how well their score is to their actual behaviour.

Question 30

How can a known-groups paradigm be used to measure criterion validity?

Answer 30

A known groups paradigm allows researchers to see whether scores on the measure can discriminate among 2 or more groups whose behaviour is already confirmed. It can also be used to validate self-report measures.

Question 31

What validities measure whether there is a meaningful pattern of similarities and differences among related measures?

Answer 31

Convergent and discriminant validity

Question 32

An example of convergent and discriminant validity is…

Answer 32

Whether a self-report measure correlated more strongly with measures of similar constructs than dissimilar constructs

Question 33

What is the difference between convergent and discriminant validity?

Answer 33

Convergent sees whether a measure of a construct is the similar to another measure of the same/similar construct - while discriminant is whether a measure of a construct is similar to a different measure of a different construct. It should also differentiate between similar symptoms but different diagnoses. Discriminant is often for near neighbours (similar but fundamentally different constructs)

Question 34

Reliability and validity

Answer 34

They are not the same - values may be less valid than reliable but it can not be more valid than it is reliable. Reliability is nescessary for validity.