研一上學(xué)期研究方法mbmz01lecture

1、MBMZ01 Research MethodologyChapter 3Measurement Foundations:Validity and ValidationConstruct DefinitionsConstruct validity refers to the degree to which inferences can legitimately be made from the operationalizations in your study to the theoretical constructs on which those operationalizations wer

2、e based. (Source: Knowledge Base)There is no direct way to measure constructs, because they are conceptual phenomena. As a consequence, there is no direct way to assess construct validity. It must be inferred from a variety of criteria. These inferences are made by examining scores on measures and c

3、omparing them with theoretical propositions about how the scores should perform. Construct valid measurement results in a close correspondence between the construct of interest and the scores provided by the measures.The most useful conceptual definitions have two elements:Construct domainNomologica

4、l networkConstruct DefinitionsIdentifies the nature of the construct by specifying its meaning.Explains what a researcher has in mind for the constructContains a dictionary-like statement that describes the construct domain. Speaks to what is included in the construct.Construct DomainNomological Net

5、workSpecifies how values of the construct should differ across cases and condition.Relationships between a construct under measurement consideration and other constructs. This network would include the theoretical framework for what you are trying to measure, an empirical framework for how you are g

6、oing to measure it, and specification of the linkages among and between these two frameworks.Construct Definition IllustrationPersonal computer satisfaction is an emotional response resulting from an evaluation of the speed, durability, and initial price, but not the appearance of a personal compute

7、r. This evaluation is expected to depend on variation in the actual characteristics of the computer (e.g. speed) and on the expectations a participant has about those characteristics. When characteristics meet or exceed expectations, the evaluation is expected to be positive (satisfaction). When cha

8、racteristics do not come to expectations, the evaluation is expected to be negative (dissatisfaction). People with more education will have higher expectations and hence lower computer satisfaction than those with less education.Construct Definition IllustrationComputer satisfaction as defined refer

9、s to speed, durability, and price.It also expected that satisfaction with appearance is NOT part of the definition.Hypothetical computer satisfaction questionnaireMy satisfaction with:Very Dissatisfied DissatisfiedNeither Satisfied nor DissatisfiedSatisfiedVery Satisfied1. Initial price of the compu

10、ter123452. What I paid for the computer123453. How quickly the computer performs calculations123454. How fast the computer runs programs123455. Helpfulness of the salesperson123456. How I was treated when I bought the computer12345Multidimensional ConstructsOne-dimensional constructsE.g. Computer sa

11、tisfactionMultidimensional constructsSatisfaction with computer speedSatisfaction with computer durabilitySatisfaction with computer priceLecture 2Conceptual to Operational RelationshipsConstruct validityPresent when there is a high correspondence between the scores obtained on a measure and the men

12、tal definition of the construct it is designed to represent. (Line b1 and b2)Construct validation Methods used to estimate a measures construct validity.Construct Validity ChallengesTwo major challenges confront construct validity.Random errorsCompletely unsystematic variation in scoresSystematic er

13、rorsIn measurement, when scores from a measure consistently vary from construct validity because of contamination and/or deficiency. Random ErrorsTaking an average of several items designed to measure the same thing is one way to address random errors.Random errors tend to “average-out” across multi

14、ple items.The more items, the more successfully this type of random error is controlled.Systematic ErrorsContaminationIn measurement, the portion of scores that measures something other than the defined construct. (多測(cè))DeficiencyIn measurement, the portion of the defined construct that is not capture

15、d by scores on a measure of the construct. (少測(cè))Items 1 and 2 relate to satisfaction with price.Items 3 and 4 related to satisfaction with speed.Items 5 and 6 ask about satisfaction with the purchasing experience. (contamination)The measure is deficient because there are no items capturing satisfacti

16、on with computer durability.My satisfaction with:Very Dissatisfied DissatisfiedNeither Satisfied nor DissatisfiedSatisfiedVery Satisfied1. Initial price of the computer123452. What I paid for the computer123453. How quickly the computer performs calculations123454. How fast the computer runs program

17、s123455. Helpfulness of the salesperson123456. How I was treated when I bought the computer12345Construct ValidationContent ValidityReliabilityConvergent ValidityDiscriminant ValidityCriterion-Related ValidityInvestigating Nomological NetworksContent ValidityA measure is content valid when its items

18、 are judged to accurately reflect the domain of the construct as defined conceptually.Content validation ordinarily has experts in the subject matter of interest provide assessments of content validity.Contamination? Deficiency?Items measuring crossover between workplace and non-workplace activities

19、4我會(huì)在工作時(shí)間內(nèi)以即時(shí)通訊處理私人或社交事務(wù)。123455我會(huì)在工作時(shí)間內(nèi)以即時(shí)通訊處理與工作有關(guān)的事務(wù)。123456我會(huì)在工作時(shí)間內(nèi)回覆與私人或社交有關(guān)的即時(shí)通訊。123457我會(huì)在工作時(shí)間內(nèi)回覆與工作有關(guān)的即時(shí)通訊。123458在一天當(dāng)中，我控制著是否將工作與社交活動(dòng)結(jié)合在一起。12345ReliabilityRefers to the systematic or consistent variance of a measure.Proportion of observed score variability (circle on the far right) that is overl

20、apped by the middle circle (systematic) representing true score variability. Types of ReliabilityInternal consistency reliabilityThe similarity of item scores obtained on a measure that has multiple items. It can be assessed when items are intended to measure a single construct.Inter-rater reliabili

21、tyThe degree to which a group of observers or raters provide consistent evaluation.Stability reliabilityThe consistency of measurement result across time.Reliability and Construct ValidityReliability speaks only to a measures freedom from random errors. It does not address systematic errors involvin

22、g contamination or deficiency.Reliability is necessary for construct validity but not sufficient. It is necessary because unreliable variance must be construct invalid. It is not sufficient because systematic variance may be contaminated and because reliability simply does not account for deficiency

23、.Reliability addresses only whether scores are consistent; it does not address whether scores capture a particular construct as defined conceptually. Convergent ValidityIs present when there is a high correspondence between scores from two different measures of the same construct.The area crossed wi

24、th vertical lines shows the proportion of variance in scores from the two measures that is convergent.The area crosses with horizontal lines shows common construct valid variance.The area covered only by vertical lines shows where the two measures share variance that represents contamination from a

25、construct validity perspective. Convergent validity also does not address whether measures are deficient.Discriminant Validityis present when measures of constructs that are supposed to be independent are found to have a low correspondence.ExampleIf scores on the researchers measure of computer sati

26、sfaction show little or no relationship with scores from a measure of satisfaction with computer appearance, then discriminant validity evidence is obtained.Convergent Validity and Discriminant Validity收斂效度指一份測(cè)驗(yàn)分?jǐn)?shù)要能夠和其他測(cè)量相同理論建構(gòu)或潛在特質(zhì)的測(cè)驗(yàn)分?jǐn)?shù)間具有高相關(guān)區(qū)別效度指一份測(cè)驗(yàn)分?jǐn)?shù)要能夠和其他測(cè)量不同理論建構(gòu)或潛在特質(zhì)的測(cè)驗(yàn)分?jǐn)?shù)間具有低相關(guān)Criterion-Related ValidityIs present when scores on a measure are related to scores on another measure that better reflects the construct of interest.No