Types of measures (operational definitions)
self-report, observational, physiological
define self-report measure
participant’s answers to questions about themselves (questionnaire or interview)
define observational measure
recording observable behaviors (i.e. allergies - how often someone sneezes)
define physiological measure
recording biological data. (i.e. EEG/ERP, HR, hormone levels)
two types of measurements
categorical, quantitative
define categorical variable
nominal. levels are qualitatively distinct (i.e. species, nationality)
define quantitative variable
meaningful numbers (i.e. height, weight, scales)
includes ordinal, interval, ratio
define ordinal variable
a type of quantitative variable. numbers represent a rank order. distance between numbers don’t have meaning.
ex) order of racers
define interval variable
a type of quantitative variable. numbers represent equal distances, no true zero.
ex) IQ, shoe size, pain scale
define ratio variable
a type of quantitative variable. numbers represent equal distances, zero represents none of the variable being measured.
ex) number of correct answers. episodes watched, height
3 types of reliability
test-retest, inter-rater, internal
define test-retest reliability
consistent scores, regardless of time
define inter-rater reliability
consistent scores, regardless of researcher measuring
define internal reliability
consistent scores, regardless of how researcher asks the question (i.e. different depression tests, same results)
2 ways to evaluate reliability
scatterplot (visual), correlation coefficient ‘r’ (statistical)
define Cronbach’s alpha
statistic used to evaluate internal reliability (similar to r-coefficient)
define validity
accuracy of measurement
is the operationalization measuring what it’s intended to measure?
face, content, convergent, discriminant
define face validity
it looks like what you want to measure
define content validity
the measure contains all parts your theory says it should contain
define criterion validity
is the measure related to a concrete outcome that it should be related to? can be assessed with the known-groups paradigm
define known-groups paradigm
does the measure align with a set of groups whose behavior is already well understood
ex) testing a depression scale by giving it to individuals who are already diagnosed as depressed
define convergent validity
a measure should strongly correlate with other measures of the same construction
ex) comparing one depression scale to another
define discriminant validity
a measure should correlate less strongly with measures of different constructs
ex) comparing a depression scale to a physical health scale
reliability vs. validity
reliability: how well a measure correlates with itself
validity: how well a measure is associated with something else
reliability is necessary for validity
