Textbook Notes (363,559)
Canada (158,426)
Psychology (9,578)
PSYB01H3 (585)
Anna Nagy (283)
Chapter 5


4 Pages
Unlock Document

University of Toronto Scarborough
Anna Nagy

6thOCTOBER, 2013 PSYB01CHAPTER 5 MEASUREMENTCONCEPTS SELF-REPORT TESTS AND MEASURES Detailed research on reliability and validity is often carried out on self-report measures of personality and individual differences  o Often use self-report measures to study psychological attributes e.g. IQ, abilities, potential o Some aimed at basic personality processes  Measures are used to make better clinical diagnoses, career-choice decisions, hiring decisions  It is wise to use existing measures of psychological characteristics rather than develop your own o Be able to compare your findings with prior research that uses the measure RELIABILITY OF MEASURES  Reliability - consistency or stability of a measure of behavior o A reliable measure does not fluctuate - yield the same result each time you administer a test on the same person e.g. IQ test  Fluctuate = error exists  Any measure comprises of 2 components: (1) truescore - the person's real score (2) measurementerror - the extent that a measure is unreliable  A reliable test will have a diagram of scores that are concentrated in the middle of a normal distribution. Thus if measure is reliable, single administration should closely reflect true score  A unreliable test will have a more spread out scores in a normal distribution. Study behavior using such tests will be a waste of time, conclusions are meaningless, results can't be repeated  Reliability is achieved by: o Careful measurement procedures o Careful training observers o In many areas, multiple observations of the same variable e.g. # of items increased To know how reliable a measure is, we assess the stability of measures using correlation coefficients  o Pearsonproduct-momentcorrelationcoefficir ) tells us how strongly 2 variables are related, ranging from -1 to +1  When talking about reliability, correlation usually is called a reliabilitycoefficient o We need to obtain at least 2 scores on the measure from many individuals  If reliable: 2 scores should be very similar Test-retest reliability  Test-retestreliabilityis assessed by measuring the same individuals at 2 points in time o 2 sets of data gotten from the same group of people at 2 different times, then correlation of the 2 sets of data is calculated. If they have very similar scores, then measure reflects true score o Reliability coefficient should be at least 0.80 for a measure to be reliable  Problems: o High correlation is artificial, because people remember how they responded  Solved by usiAlternateformsreliabilitynistering 2 different forms of the same test at 2 points in time o Some variables change from one test period to the next e.g. mood state o Obtaining 2 measures from same people at 2 points in time may be difficult sometimes 1 | P a g e th 6 OCTOBER, 2013 Internal consistency reliability  It is possible to assess reliability at only one point in time: each item as repeated attempts to measure a concept o Respond similarly across these repeated items = reliable  Internalconsistencyreliability assesses how well a certain set of items related to each other o Since items measure same variable, they SHOULD yield consistent results o One indicator is a value calledch'salpha, which calculates how well each item relates to every other item  Produces interitemcorrelations  Based on average of all interitem correlation and # of items  Very informative o Useful when: Items that do not correlated with others can be removed to increase reliability   Need to construct a brief version of a measure Interrater reliability  In some research, raters observe behaviors and make ratings or judgments o Single observations of one rater might be unreliable o Solution to this is to use at least 2 raters to who observe the same behavior o Interraterreliabi- extent to which raters agree in their observations  A common indicator = Cohen'skappa Reliability and Accuracy of measures Reliability tells us about error in the measurement but it does not tell us about whether our measure accurately reflects the variable of interest  e.g. foot-based measure is validndicator of intelligence o This leads us to a consideration of the validity of measures VALIDITY OF MEASURES Second requirement of a quality operational definition   From CHAPTER 4, internalvalidityflects degree to which a claim of causation can be drawn from an experiment  Constructvalidity – the adequacy of the operational definition of variables o Degree to which the operational definition of a variable reflects the true theoretical meaning of the variable E.g. shyness: should be accurate indicator of that trait o Whether the measurement actually measures the construct it is intended to measure (in terms of measurement)  E.g. GRE supposed to predict one’s success in grad school: validity determined by whether it does measure this ability to success o “Construct” denotes a variable that is abstract and needs an operational definition Indicators of Construct validity Facevalidity–the measure appears “on the face of it” to measure what it is supposed to measure  o The simplest way to argue a measure is valid o Not sophisticated, only a judgment based on appearance o PCL-R appears to assess psychop
More Less

Related notes for PSYB01H3

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.