Measurement Flashcards
measurement
assigning numbers or categories to individuals in your sample
self-report
a measure in which participants report on their thoughts feelings behaviors in a survey or interview
behavioral
a measure in which some aspect of the participants behavior is observed and recorded
- asking participants to memorize word, do math, then recall words
-participants do behavioral task
physiological
a measurement that involves recording a physiological variable
- fMRI
nominal
name labels, no numbers involved
ordinal
ordered scores
first, second, third
interval
equal units, no true zero
- temperature
- 0 degrees does not absence of heat
ratio
equal units, meaningful zero
- time
- 0 seconds means no time
reliability
how consistent is your measure?
validity
are you really measuring what you think you’re measuring?
characteristics of reliability
time, observers, items
test-retest reliability
reliability across time
- if you measure the same person again, do they get a similar score?
inter-rater reliability
reliability across observers
- do different observers give the same subject the same score?
internal reliability
reliability across items
- if you ask the same question in a different way, do they give a similar answer?
measurement validity
does the measurement process measure what it is intended to measure?
construct validity
how accurately does the operational definition capture the construct?
how do we know if a measurement is valid?
subjective tests
empirical test
subjective tests
is the content of the measure appropriate for the construct?
empirical tests
is the measure related to other measures in a way that makes sense?
face validity
measurement seems to capture the intended variable
does the test appear to test what it aims to test
content validity
measurement covers all aspects if the construct that it is supposed to measure
- does the test reflect all aspects of the conceptual definition of the construct?
criterion validity
measurement correlates with a relevant behavioral outcome
- if we compare measurement to irl, does it correlate?
- assessing measurement against real world situations
convergent validity
measurement correlates with measures of similar constructs
example:
developing depression scale
- scale has to correlate with depression symptoms
convergent validity helps new studies to come to a conclusion with existing scales
discriminant validity
measures doesnt correlate with measures of dissimilar constructs
- looking for things with no relationship
- can’t get out of bed in the morning
- doesn’t mean they necessarily they have depression, it might be because they have a health condition