Lecture 5: Measurement Flashcards
(24 cards)
construct validity
ask whether the measure of the variable actually measure what they are supposed to
conceptual variable
- variable at an abstract level
- also called a construct
operational variable
-variable as it is specifically and concretely measured or manipulated
self report measure
-written or verbal responses to questionnaire or interview questions
observational measure
- observing and recoding behavior (behavioral measure)
- includes tests, physical traces of behavior, records
physiological measure
- recoding biological data
- e.g. brain activity during happy picture stimuli
operational variables coded with 4 different scales of measurement
- nominal
- ordinal
- interval
- ratio
nominal scale
- level are categories with different label/names
- cannot be ordered in anyway
- levels represent difference in kind, but not quantitative information
Example of a nominal scale
- sex
- color
- major
ordinal scale
- levels are categories with different names/labels (perhaps numerals)
- can be ranked ordered from low to hight
- can not determine size of difference between two levels
ordinal scale example
- class standing
- hotness of chilies
- ranking of music from least to most favorite
interval scale
- levels are different numerical values
- equal size intervals between levels
- no true zero point
- can determine direction and size of difference
interval scale example
- temperature (C or F)
- IQ
- questionnaire scales
ratio scale
-levels are different numerical values
-equal size intervals between levels
-true zero point
can determine direction, size of difference, and describe differences in terms of a ratio
example of ratio scale
- weight
- height
- temp in K
categorical variable
- levels are categories with different labels/names
- nominal variables
quantitative variables
- levels are coded with meaningful numbers
- interval and ratio variable
note on classification of ordinal variables
- book argues ordinal variables are quantitative since there is some quantitative information
- many sources however classify them as categorical since they are often labeled with categorial labels such as mild medium and hot
Questionnaire Scales classification
-approximately interval scales
construct validity
-need to question reliability and validity
reliability
- is our measurement consistent across time, raters, items
- does the measure correlate with itself
validity
-is our accurately measuring what it should be measuring
test retest reliably
- consistent score across time for same individual
ex: IQ test
interrater reliability
88
