Understanding Theoretical Constructs and Operational Definitions
School
University of Regina**We aren't endorsed by this school
Course
PSYC 204
Subject
Psychology
Date
Dec 10, 2024
Pages
7
Uploaded by ProfProton23622
Theoretical constructs VS Operational Definitions-Constructs are abstracts, not physical. Our confidence-We need operational definitions because TC.-Which are definitions of constructsoPhysiologicaloSelf-reportoBehavioral observationoAny type of measurement of a theoretical construct.oEx. Wealth (Theoretical construct because it’s an abstract.) Operational definition (What is wealth? Having good health? Or money?)-Diener’s 5 item subjective well-being scale-The perception of how they view happiness and rate that perception through a 5 item subjective ona 7 point scale.-1 being strongly disagree-7 being strongly agree-10-point ladder of life scale (Cantril)o0 being the bottom of the ladder and 10 being the top.-Kahneman and Deaton 2010-What about emotion? What about moments of feeling possitve affect?oAsking more directly questions about the measure of happiness-Ladder and Positive affect Chart – Happiness strongly related to income-Positive affect has an increase of happiness as income increases however it stops the more income one gets.-Life ladder evaluation constantly increases as income increases-Do we define happiness as success?-Or hapiness our emotions?-Operational definitions fails if:oThey ignore important elements of a construcLacks important elements.Doesnt focus in all aspects of a constructoThey include irrelevant elements
Goes even further from a certain construct making it more complex rather than simple.Reliability and Validity-Core issues in measurements-Reliability is we’re measuring constructs constantly -The extent to which operational definitions are highly repeatedo3 types of reliabilityTest-retestExtent to which scores from the same “instrument” or type of measurement are similar across 2 measurement periods. (they remain highly consistant)Ex. Weighing in January 93KG and weighing again in February with 92 KG.Inter-raterThe extent to which the scores from two or more observers are similarObservations from observers. - judgesThey have to decide what determines a measurement prior to observation in order to get similar scores / high inter-rater.Examined by scatterplotInternalThe extent to which answers to items within a single measure are associated with answers to other itemsConsistent responses to all types of items within the measurement.-Cronbach’s Alpha StatisticoLooks at the correlation of the items of internal reliabilityo0.7 and above meaning questions are too similar and relate.Validity is about if we’re measuring what we want to
-Construct validity is the extent to which an operational definitions measures what is it supposed to measure.-Will you go off topic?-Face ValidityoThe extent a measure looks like what you want to measureoEx. IQ test looking like it measures IQ. = high validity.oIQ test looking off topic (IQ) = low validity-Content validityoA measure contains all parts that a theory says it should containoThe type of validity that covers all important material from the construct.oEx. PSYC 204 final exam covering most relevant information from the course = high validityoThe exam lacking important material like reliability and validity = lowvalidity.-Criterion ValidityoThe extent to which scores on the measure are associated with concrete behavioral outcomes they should be associated with.oEx. A test testing happiness. A happy person scoring high on the measurement. oEX. LSAT. Admission grad tests. Does the test measures students’ knowledge about the law? If so, high grades will mean they will do good in law school; however, if students score low, will it mean that students are less knowledgable about the law? Or does the test measure what its supposed to measureoSelf-report is importantoThree types:PredictiveThe extent a measure accurately predicts behavior in the future.Concurrent (at the same time)The extent a measure accurately predicts current behavior
The correlation of the construct and the behavior = high validityKnown Group The extent to which scores of a measure can differentiate between groups whose behavior is already confirmed.We know these groups have been assesedEx. Professional athletes scoring high on measures of athleticism because they are professionalsIf score low it means its low known group validity meaning the measure isn’t the best way to define an athlete or they arent professional at all.oConvergent ValidityThe extent to which scores on a measure correlates with other measures of the same construct and measures of similar constructs.Ex. Passing my French quiz in uni vs in duolingoTwo different measures (uni and duolingo) of the same construct (French)oDiscriminant or Divergent ValidityThe extent to which a measure does not correlate with things it should not correlate with.Ex. -A measure must be reliable first in order to be valid-However, a measure can be reliable but not valid.oEx. Constant scores most of the time makes it reliable however what makes it unvalid is the fact that it isnt true or its wrongTypes of measuresSelf-report-It’s just asking participants about information
-Questionnaires-Interviews-Rating scales-Assume people: (TRUST)oHave access to their psychological processesoWon't lie (Good sense of behavior)oAre not influenced by contextSocial desirability biasoThey are the only way to measure many constructs despite their problems.oHaving to ask a participant what they dream of would require the use of a self-report because only the individual can see his dreams and we can’t prove he is lying.oTwo types of self-report:RetrospectiveAsking participants things theyve done in the past Might be invalid bc we dont know if theyre right or notProspective reportSimilar problemParticipants speculate about future behaviorsWhat the participants say they will do and doesnt end up doing anythingPhysiological-Operationalizes a variable by recording a biological data-Like the use of MRI, heart rate, blood pressure, blood test, brain, polygraph-Highly reliable (Very objective)-Require specialized equipment-Can be intrusive – taking ppls blood for examination-Questionable validity despite being reliableoPolygraph, lying detector oIs it really that people are lying or are there other factors like stress? Or nervousness?oNot entirely accurate
-Physiological measures validity have to be stablished by self-report or observational measures.oEx. Fmri checks an individuals part of brain is light up when an individual is happy. But in order to make the operational definition weneed to measure the participant if he was happy.Observational (Behavioural)-FrequencyoCount number of occurrences for a specific periodoStuttering. Participants being asked to read a certain amount of pages and measuring how much times people would stutter between a specific amount of pages.-LatencyoMeasure amount of time it takes to participants to respond to some type of stimulusoAny measure of reaction time is latency-Number of errorsoHaving what counts as correct, we can count the number of errosoEx. Implementing an knowledge exam about a specific topic. The amount of errors would be the measure of the knowledge of the topic.-Strong validity depending on the constructoNot influenced by some of the problems with retrospective and prospective self-report measuresoIt’s more valid due to the fact that self-report measure can be biased and observational measures are done by other people than yourself.oEx. Cell phone usage. On a self report measure people would say that they dont use their phone as often in class. However, we do not know that unless cameras or observers confirm the exact amount of times one went on their phone. So it could either be more or exact.-Less influenced by response biasesoThey sort of prove the response biasesoDetermine what they see-Will be influenced by situational factors
oKnowing that observational measurements are occuring can make it invalid due to the participants knowledge of being observed. Which makes them biased and behave.-Don't necessarily address the underlying cause of the behavioroWhy it happened? What led to the cause of that behavior? Was it that there were important things were happening? OR was it just for fun?-Not possible for some variablesoWhere self-report measure kick inoEx. Dreams. We cant observe people’s dreams the only weay to know they is through self-report measures. Which operalization is best?Donald Campbell-No one measures perfect-Convergence validity-The use of two measures to determine a construct for better validity-There is no right or wrong measurement all are the same to measure.MultipleMeasurementModalities