Chapter 5 - Measurement

Question 1

Which variables are measured in non-experimental vs. experimental designs?

Answer 1

• non-experimental: all variables measures

- experimental: DV measured only

Question 2

types of scales

Answer 2

• qualitative variable

- quantitative variable

Question 3

qualitative variable

Answer 3

• things with no numerical values (words, concepts, communication, etc.)
• ex. Nominal scales

Question 4

nominal scales

Answer 4

assigned numerical values are meaningless (ex. Gender, sexuality, religious affiliation)

Question 5

quantitative variable

Answer 5

• things with meaningful numbers that can be placed on numerical scales
• ex. Ordinal scale, Interval scale, Ratio scales

Question 6

ordinal scales

Answer 6

when things have a meaningful order (ex. Socioeconomic status, ranking a set of pictures)

Question 7

interval scales

Answer 7

variables indicate order and the difference between each number is equivalent (ex. Temperature)

Question 8

ratio scales

Answer 8

have all the characteristics of ordinal and interval scales, but the 0 is meaningful because it indicates a complete absence of something, so you can then take the difference between numbers and create a ratio (ex. Using Kelvin to measure energy – 0 degrees Kelvin is an asolute absence of all heat/energy… it’s equivalent to ~-275 degrees celcius)

Question 9

4 things to consider when constructing a measure

Answer 9

• Reliability: Does it measure the construct with little error? Is it a stable measure? (True score + error = person’s combined score. We want to minize the effect of the error)
• Construct validity: Are we measuring what we think we’re measuring?
• Internal validity: Can we infer causality?
• External Validity: Can we generalize our findings beyond this group and setting?

Question 10

3 types of reliability

Answer 10

• test-retest reliability
• inter-rater reliability
• internal consistency reliability

Question 11

test-retest reliability

Answer 11

• how consistent is the measure across time?
• Evaluated using the Pearson correlation coefficient (r)
• The largest positive correlation indicates a higher test-retest reliability

Question 12

inter-rater reliability

Answer 12

• how consistent is the measure when different people are rating it?
• Evaluated using the Intraclass correlation coefficient (ICC) and/or Cohen’s Kappa
• The ICC only goes from 0-1 (no negative values) with 1 being complete agreement and 0 being no agreement
• Other uses of inter-rater reliability: Behavioural coding, thematic/content coding, used when you ask an open-ended question to figure out the overarching themes in your responses, archival research, personalty measures

Question 13

internal consistency reliability

Answer 13

• Are scores similar across different questions?
• Evaluated using Cronbach’s alpha (a) combining inter-item correlations
• 0.9 is ideal, between 0.6 to 0.9 is good, less than 0.6 isn’t very good
• Evaluated using item-total correlation
• Used on multiple choice exams

Question 14

construct validity

Answer 14

• Does content of measure reflect the meaning of the construct? (ex. face validity, content validity)
• How does this measure relate to other measures and behaviours? (ex. predictive validity, concurrent validity, convergent validity, discriminant validity)

Question 15

face validity

Answer 15

• Look at each item
• Does it look like it’s assessing risk-taking?
• Usually happens, not not a requirement of measures

Question 16

content validity

Answer 16

• Look at the whole measure

- Is it capturing all the important parts of what it means to be a risk taker in adolescence?

Question 17

predictive validity

Answer 17

• Predicts future, conceptually related behaviours

- Do people with high scores on your measure at T1 go on to do relavant behaviours at T2?

Question 18

concurrent validity

Answer 18

• Related to relvant outcomes assessed at the same time
• Do people with high scores on your measure behave in ways you’d expect them to behave if they were high on this construct?
• Emphasis on behaviours

Question 19

convergent validity

Answer 19

• Related to scores on other measures of the same construct
• Do people with high scores on your measure have high scores on theoretically similar measures of the same construct? (positive correlation)
• Emphasis on feelings/psychometric properties

Question 20

discriminant validity

Answer 20

• Not related (ie. Zero correlation) to what it shouldn’t be related to
• Do people with high scores on your measure randomly vary in how much they show constructs that could be alternative explanations of what your scale is measuring?

Question 21

2 aspects of measures

Answer 21

• True score: person’s real score on the variable

- Measurement error: the more error there is, the less reliable the measurement is

Question 22

reactivity

Answer 22

what the person is like when he or she is aware of being observed