well-being is a ___ construct. to measure well-being, we need the process of _____. what we actually measure are called ____.
latent
operationalization
indicators
what is the criteria for measurement formula and what do the variables represent?
X = T ± E
X represents your score
T represents your true happiness
E represents measurement error
name and define the two types of measurement error
random measurement error: random fluctuations in score that influence score in unpredictable ways. unrelated to construct of interest.
systematic measurement error: non-random fluctuations in score that influence score in predictable ways. related to construct of interest (ex. social desirability bias).
define reliability
- are we consistently measuring some construct?
- how much random measurement error there is in an individual’s score.
describe the following:
1) split-half reliability
2) test-retest reliability
3) inter-rater reliability
4) internal consistency
1) split-half: randomly create 2 subsets from a single measure and correlate average score of subset 1 with average score of subset 2.
2) administer measure at different time points and find correlations between time points (ex. average at time 1 and average at time 2).
3) different people make simultaneous ratings of the same target(s) and find agreement across raters (instead of correlations it’s kappa).
4) using multiple items in the measure, look at how items correlate with each other (averaging ALL correlations, cronbach’s alpha).
define validity
- are we measuring what we’re supposed to?
- how much an individual’s score reflects their true score and systematic measurement error.
describe the following:
1) content validity
2) construct validity
3) criterion validity
4) internal validity
5) external validity
1) content: aka ‘face validity.’ ex. look at questions themselves and decide if they’re measuring happiness.
2) construct: does operationalization match the theory?
3) criterion: encompasses convergent, discriminant, and predictive validity.
4) internal: is it my independent variable(s) causing the changes in my dependent variable(s)?
5) external: are my results generalizable?
what are the pros (1) and cons (2) of closed-ended responses in terms of self reports?
pros:
- time and cost-efficient.
cons:
- do not capture underlying dynamics.
- # of responses, order, and labels influence reliability and validity.
what are the pros (1) and cons (2) of open-ended responses in terms of questionnaires?
pros:
- captures underlying dynamics.
cons:
- interpretations may be subject to bias.
- analysis is labour intensive, time consuming and expensive.
partial solution? use computer-analysis software.
define satisficing + why do we do it?
satisficing: tendency to look for cues in the question/response options that point to a reasonable answer.
why? want to conserve cognitive effort/do not have sufficient cognitive resources or skills to think about an accurate answer.
define acquiescence + why do we do it?
acquiescence: tendency to agree with an assertion made in a question/statement regardless of the content.
why? to be polite/avoid confrontation/deference to higher social status individuals/satisficing.
define social desirability bias + why do we do it?
social desirability bias: tendency to choose responses that allow one to present oneself in a positive manner to others.
why? to maximize pleasure and minimize pain/other-deception + self-deception/protects self-esteem.
what’s an example of order effects in terms of questionnaires?
- evaluations of later questions/statements are influenced by evaluations of earlier questions/statements.
- ex. “how often are you dating?” followed by “how happy are you?”
- self-perception theory: you construct your present attitudes based on past behaviour and then make inferences about yourself.
what are the three components of subjective well-being (SWB)?
1) people’s long-term levels of positive affect (PA),
2) lack of negative affect (NA) and,
3) life satisfaction.
describe bradburn’s affect balance scale and its limitations (2)
- answer yes/no to two sets of questions (one PA-oriented, the other NA-oriented).
- affect balance = TOTAL - TOTAL.
limitations:
- too few items.
- too narrow items.
describe fordyce’s affect measure and its limitations (1)
- answer questions via scale of 1-11 and find average.
limitations:
- response options are difficult to translate (ex. what does 7 vs 8 mean?)
describe the affectometer 2 and its limitations (1)
- 40-item scale to measure affective well-being and life satisfaction with 10 underlying factors.
- measures balance of pleasant and unpleasant feelings in recent experiences.
limitations:
- discriminant validity: may reflect other individual differences (ex. self-esteem, optimism, which are NOT parts of SWB).
describe the positive and negative affect schedule (PANAS) and its limitations (2)
- measures PA and NA using 5-point rating scale (very slightly to very much).
- can use different time frames (ex. “today,” “past few days”, “past few weeks”) which captures valence and intensity.
limitations:
- recall biases and inaccuracies.
- may use mood to judge emotions and vice versa.
which is the most widely used measure of affect?
PANAS
describe cantrill’s life satisfaction ladder
- ladder scale from 0-10, participants asked to point where on the ladder they stand.
describe the satisfaction with life scale (SWLS) and its limitations (1)
- answer questions from strongly disagree - strongly agree.
limitations:
- every item is positively worded (acquiescence hazard).
describe the average domain satisfaction measure and its limitations (1)
- rate satisfaction in a list of specific life domains.
limitations:
-which domains are relevant to everyone?
people may weigh domains differently.
in conclusion, which measures are “better” (affect measures or life satisfaction measures)? provide evidence.
- life satisfaction measures!!!
- internal consistency (cronbach’s α = .8+)
- consistency across short time intervals and situations (ex. r = .8+) and reliability lower over long-term, single-item measures (able to measure true changes)
- about 74% reliable variance
- good convergent/discriminant/criterion/ predictive validity too
what are the six dimensions of psychological well-being?
1) self-acceptance
2) positive relations with others
3) environmental mastery
4) purpose in life
5) personal growth
6) autonomy
