PSYC 200W Lecture Notes - Lecture 6: Concurrent Validity, Intraclass Correlation, Construct Validity
28 Jun 2018
School
Department
Course
Professor
Observational measures - direct observation of behavior
Physiological measures - studies relationship between bodily processes and behavior
Self-report measures - replies people to give to questionnaires and interviews
Cognitive self-reports: measure what people think about something
Affective self-reports: participants' responses regarding how they feel
Behavioral self-reports: participants' reports of how they act
Psychometrics - study of psychological measurements
Converging operations/triangulation - when different kinds of measures provide the same results,
we have more confidence in their validity
Scales of measurement:
Nominal scale: numbers assigned to participants' behavior (labels)
Ordinal scale: rank ordering of a set of scores, however, cannot measure difference
between two participants
Interval scale: equal differences between numbers reflect equal differences between
participants on the characteristic being measured. Does not have true zero value -
temperature because 0 degrees does not mean absence of temperature, latitude
Ratio: true zero value, weight - 0 weight means absence of weight
Reliability - consistency/dependability of a measure
Observed score = True Score + Measurement Error
True score - score that the participant would have obtained if our measure was perfect without any
error
Measurement error ----> Error variance
Participant
oTransient factors (temporary) - They are in a bad mood, health, anxiety, fatigue
oStable attributes - Intelligence, ability, general motivation
Not Participant
oSituational factors - room temperature, lighting, researcher demeanor
oMeasure factors - ambiguous qs, too long, induce fear
oMistakes - data entry errors, computer glitch
Measurement error undermines the reliability of the measures used. Negative correlation
Total variance = variance due to true score + variance due to measurement error
Systematic variance: Total variance in a set of scores that is associated with participants'
true scores
Error variance - variance due to measurement error
Reliability = true score variance/total variance
Estimates usually obtained via correlations between measures of same attribute
Correlation coefficient - r - can range from 0 to 1
Sufficient reliability >= .70
Types of Reliability:
Test-retest - consistency of participants' responses on a measure over time
Interitem - degree of consistency among scale items - also called internal consistency
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
Physiological measures - studies relationship between bodily processes and behavior. Self-report measures - replies people to give to questionnaires and interviews. Cognitive self-reports: measure what people think about something. Affective self-reports: participants" responses regarding how they feel. Behavioral self-reports: participants" reports of how they act. Converging operations/triangulation - when different kinds of measures provide the same results, we have more confidence in their validity. Nominal scale: numbers assigned to participants" behavior (labels) Ordinal scale: rank ordering of a set of scores, however, cannot measure difference between two participants. Interval scale: equal differences between numbers reflect equal differences between participants on the characteristic being measured. Does not have true zero value - temperature because 0 degrees does not mean absence of temperature, latitude. Ratio: true zero value, weight - 0 weight means absence of weight. Observed score = true score + measurement error. True score - score that the participant would have obtained if our measure was perfect without any error.
