Session 3 : Psychometrics Flashcards
Psychological measurement
the assignment of scores to individuals so that the scores represent some characteristic of the individuals (some constructs are not observable)
Four steps in the measurement process
- conceptually defining the construct
- operationally defining the construct
- implementing the measure
- evaluating the measure
Conceptual definiton (conceptually defining the construct)
a conceptual definition of a psychological contruct describes the behaviors and internal processes that make up that construct, along with how it related to other variables
Operational definiton (operationally defining the construct)
a definitions of a variable of precisely how it is to be measured
Three categories of operational definition
- self-report measures
- behavioral measures
- pyhsiological measures
Levels of measurement
- nominal level (assigning scores that are category levels)
- ordinal level (assigning scores that represent rank order of individuals)
- interval level (assigning scores using numerical scales in which intervals have the same interpretation throuhgout)
- ratio level (assinging scors in such way that there is a true zeor point that represents the complete absence of the quantity)
Evaluating a measure (two distinct criteria)
- Reliability
- Validity
Reliability
refers to the consistency of a measure across time, across items, and scorss researchers
Validity
the extent to which some scores from a measure actually represent the varibale they are intended to
Three types of reliability
- Test-retes reliabiliy: reliability over time
- Interrater reliability: reliability across different resaerchers
- Internal consistency: reliability across items
Test-retest reliability
the extent to which the scores obtained by a measure of a construct are consistent over time
Interrater reliability
the extent to which different observers are consistent in their judgements
Internal consistency
consistency of poeple’s responses across the items on a multiple-item measure
- observed score = true score + measurement error
- all items should reflect the same underlying construct
Internal consistency
can assess internal consistency by split-half correlation or Cronbach’s alpha
- greater than 0.7 (good internal consistency)
Different kinds of validity
- face validity
- content validity
- criterion validity
- discriminant validity
- construct validity