measurement and scale design Flashcards
what happens when there is no measurement
very dangerous
happens often in research
what are the challenges with measuring phenomenon (happiness, motivation, pain, etc)
subjective experiences
not easy to quantify
relies on participants to be truthful
what are the 3 ways participants lie about phenomenon
satisficing
- satisfactory but not truthful answer
- what “should” be answered, flip a coin, etc
social desireability
- putting best foot forward
- participant is UNAWARE they are doing it
faking good
- INTENTIONAL - conscious decision
why is measurement important for PA
to see if interventions work
can use to create attainable goals
to create guidelines
what are the two ways of measuring PA
objective assessments
subjective assessments (scales or participant report)
what are the benefits and disadvantages of objective assessments
benefits
- concrete measure (can’t lie)
- consistent and comparable between participants
disadv
- expensive, long, not as feasible
- hard to use depending on population
what are the benefits and disadvantages of subjective measures
benefits
- easy to do
- not expensive
- can use larger populations
disadv
- people can lie
how to start scale development
key informant interviews (used to find the questions to ask)
theories
expert opinion
what are the 2 types of ways people can respond to questions
categorical judgements
continuous judgements
what are categorical judgements
yes/no questions
when can categorical judgements be a problem
can be unclear
can be used for something thats not categorical (degrees of pain)
what are the 3 types of continuous judgements
direct estimation
comparative methods
econometric methods
what is direct estimation
magnitude of the behaviour / attitude
ex: scale of 1-10 happiness
what are comparative methods
seen lots in kids
harter scale
- some kids like to ride bikes adn some kids don’t, which is most like you
what is the econometric methods
two anchors on very extreme ends
- has the person estimate between the end values
what challenges are present when using scales with populations with NDD
language - needs to be appropriate for population of interest (usually gr 4)
only two options might not work
also presenting a low of info at once
timing of seasonal PA
what are the cognitive requirements needed to answer scaled questions
able to understand / interpret the question properly
recall - ability to recall past behaviours
inferences and estimation - usually overestimate rare and underestimate frequent behaviours and give easily divisible numbers (or that align with a month)
what does a reliable test mean
consistency of measure
interrater reliablity (same between two people)
test retest reliability
what does validity mean
accuracy of a measure
convergent validiy = comparing the scale to a gold standard or something it shoudl be related to
