Sem Helsinki2012

Embed Size (px)

Citation preview

  • 7/29/2019 Sem Helsinki2012

    1/18

    1

    Course Syllabus Amiram D. Vinokur

    fn: sem_Helsinki2012.doc University of Helsinki June 11-13, 2012

    Building and Testing Structural Equation Models (SEM)

    In the Social Sciences

    Course Description:

    Since the early 80's, Structural Equation Modeling (SEM) analyses, first using LISREL

    and later using EQS, AMOS and MPLUS user-friendly software packages, have gained

    prominence as they replaced the older traditional analytic methods of factor analysis and path

    analysis. SEM merges confirmatory factor analysis with path analysis and provides means for

    constructing, testing, and comparing comprehensive structural path models as well as comparing

    the goodness of fit of models and their adequacy across multiple groups (samples). Among themany advantages of SEM over the traditional path analysis is that it provides adjustment for the

    relative unreliability of the observed measures, overall goodness of fit measures and tests for

    comparing models. Without adjustment for reliability, results from traditional path analysis

    confound the substantive contribution of predictors with their relative methodological strength as

    indicated by their reliabilities. This confounding may lead to seriously flawed conclusions in the

    interpretation of the results.

    This course will cover the conceptual and technical issues relevant to the application of

    Structural Equation Modeling (SEM). Following the presentation of major conceptual issues, five

    basic structural models will be described in detail. The models vary from simple to more complex

    ones. They also cover wide range of situations including longitudinal and mediation analyses,comparisons between groups, and analyses that include data from different sources such as from

    parents, teachers, supervisors, and co-workers. The description and discussion of the models will

    provide students with the knowledge to apply SEM techniques for analyzing, evaluating and

    reporting results produced by this analytic method. This knowledge is necessary to apply SEM

    using any of the major SEM software packages.

    Prerequisites: One or more courses in statistics that included in-depth treatment of linear

    regression analysis, basic knowledge of the concepts of item analysis and internal reliability.

    Goals of the Course:

    The goal of this course is to provide students with basic knowledge of how to construct,analyze, modify, and test a variety of structural equation models and report the results of their

    analyses in a manner acceptable in refereed journals.

    The Course will be Conducted as a Lecture with powerPoint slides and with discussions.

    Class Schedule: June 11 - 13: 9am to 12noon, Lunch break, 1pm to 4pm.

    With the exception of textbooks, all of the resources for the class, including this syllabus, required

    and suggested readings, all figures, and input/output models to be shown as PowerPoint displays

    during class presentations, will be available to the students on the course workplace site at the

  • 7/29/2019 Sem Helsinki2012

    2/18

    2

    University of Michigan Ctools (https://ctools.umich.edu). Once you log in click on Resources

    (under LSA box on the left, under Announcements) and you will see the various subdirectories of

    the resources. Note, about 10 days before the class, you will received an email from me with

    information how to access the Michigan Ctools site for the course. You will also receive as

    attachment to the email the file with the powerpoint slides for the class.

    I. Course Outline (Overview of the Topics)

    Note. Enclosed in pointed parentheses are the powerpoint slides as indexed in the Table of

    content in the powerpoint file in CTOOLS website. The file name is: Sem_pp2012a.ppt

    During the lectures I will present and discuss all the major principles, topics and issues

    involved in the application of SEM to construct, test, evaluate and report path, measurement andstructural models. Most of the issues and principles will be presented and discussed with the

    presentation of a core set of five models. The models are listed below with their unique

    characteristics, and the slide number to the figure in Sem_pp2012a that represents each one.

    Model A: A model with one indicator per factor. Work-Family Conflict model (Model A

    {PP8}).

    Model B: A model with several indicators per factor, and with additional background variables

    as controls. Children's Adjustment model (Model B: {PP10}).

    Model C: A model that includes two groups to allow comparisons between the groups. Job-

    Search Intention model (Model C; {PP14a}).

    Model D: A model based on longitudinal design of data collection with a variable thatrepresents an experimental manipulation (intervention) and its mediation effects.

    JOBS II mediation Model (Model D: {PP15a}).

    Model E: A model based on longitudinal design and data collected from different sources (e.g.,

    husbands and wives). The Couples' Stress and Coping Model (Model E: {PP17b}).

    Other types of models will also be presented and discussed including path analysis models,

    measurement models, and models with reciprocal causation.

    1. What is Structural Equation Modeling (SEM)?

    a. Brief History of the roots of SEM

    b. Basics of regression {PP1a, PP1b} and factor analysis and therelationship of SEM to Path analysis and factor analysis {PP1c}

    c. The main advantages of SEM over path and factor analyses and traditional ANOVA

    and what can SEM be used for:

    1. Adjustment for unreliability of the indices {PP2a, PP2b, PP2c}

    (see Bedeian et al., 1997)

    2. Improved ability to handle multi-colinearity (compared with OLS regressions

    {PP3} (see for example Marshal and Lang, 1990).

    3. Availability of overall goodness of fit measures for the entire model.

    4. Jointly performs confirmatory factor AND structural analysis.

    5. Comparisons of models and specific paths across several groups.

    6. Modeling various types of factors (formative and reflective factors).

  • 7/29/2019 Sem Helsinki2012

    3/18

    3

    7. Conducting sensitivity analyses and various simulation analyses.

    8. Analysis of experiments/interventions that include mediator variables, that is,

    analysis of the mediation process.

    9. Using the results of meta-analyses as input to test SEM models.

    d. Basic assumptions regarding scaling, distributions, errors and causal interpretations.

    e. Choosing SEM Software: EQS, LISREL and AMOS (see Kline ch.4.6 pp. 77-85 on

    Similarities and differences).

    (For general reviews see Hoyle and Smith (1994) or Klem (2000). The article by

    Klem also includes all the terminologies and symbols used by LISREL users.)

    2. Basic concepts and notations

    For a - e: {PP4a} (See Byrne, 1994, pp 3-12)

    a. Observable (measured) one item variable, or, index (based on several observed

    variables)

    b. Observables as indicators (outcomes) of a latent factor and their factor loadings- The need to establish internal reliability (alpha)

    c. Observables as determinants of a latent factor

    d. Errors of observables and residuals as errors of latent endogenous factors

    e. Relationships among latent factors, among errors and residuals

    For f-g: {PP4b}

    f. The measurement model and the structural model

    g. Exogenous and endogenous variables and latent factors

    h. Deduced (reflective) vs. induced (formative) latent Factors

    {PP5a, PP5b, PP5c} (see Edwards and Bagozzi, 2000).

    3. Basic Concepts: the data, the estimates (model parameters)

    a. The empirical obtained correlations (S-matrix), the model implied ones (in - matrix),

    and the fit of the model to the data (based on the difference between the S and

    matrix). {PP5d}

    b. Degrees of freedom, equations and identification. {PP6a}

    c. Identification by fixing factor variance to 1 vs. fixing one of the loadings to 1.

    {PP6b, PP6c, PP6d}.

    4. The required input for SEM analysis using EQS

    a. Raw data. Raw data can be in the form of SPSS file that can be opened by EQS.b. Matrix of covariances (listwise and/or pairwise matrix)

    c. Matrix of correlations with means and standard deviation in the bottom two rows.

    5. Types of models and their specification in a diagram

    a. Simple path model: {PP7a}

    b. Latent structural model with one indicator per factor: {PP7b}

    c. Latent structural model with multiple indicators per factor: {PP7c}

    d. Non-recursive longitudinal model with reciprocal causal influence: {PP7d}

    e. Second order factor structure {PP7e}

    f. Structural model with exogenous, endogenous factors and observed variables {PP7i}

    Reading for outline sections 2-5: Ch. 5, pp. 91-122.

  • 7/29/2019 Sem Helsinki2012

    4/18

    4

    6. Illustrating the basic elements of the EQS input setup using Baggozzis model {PP7i}

    a. /Title /Specifications /Labels /Equations /Variance /Covariance

    /Matrix /Constraints /Print [Fit=all;] /Wtest /LMtest, /Endb. Some basic rules of syntax: the !!! the estimate * , dependent variable to the

    left, starting values, fixed values, unlisted covariances are fixed to = 0.

    (Information on these and other important options see the EQS manuals by Bentler, 1995

    pp. 43-82, and the User Guide by Bentler and Wu ,2002, pp.238-267).

    7. Illustrating the basic elements of the EQS output using Bagozzis Model {printout of

    Baggozzi_P.rtf can downloaded from the website section on models}

    8. Testing the measurement model of Bagozzis structural model that is shown in slide

    PP7i.

    9. Illustrating the input setup on the Work-Family Conflict model (Model A {PP8})

    (a single indicator model as in PP7b). How the relative reliabilities of the measures are

    adjusted with one measured variable (that is, with one empirical indicator).

    10. Continued: One indicator per factor model: Illustrating the output of the model using

    the output of the Work-Family Model. This will include the Basic elements of the

    EQS output (Empirical and estimated covariance matrix, residual matrix, RMR, measures

    of goodness of fit, measure of misfit, non standardized and standardized parameters, the

    measurement and structural equations, the program's output with suggestions for

    modification of the model to obtain better fit, and more). (For another example see Froneet al., 1992; another important justification for a one indicator to factor model is Liang et

    al., 1990).

    11. Model evaluation using chi-square, goodness-of-fit and misfit indices {PP8a, PP9}. The

    Chi-square/df, NFI, NNFI (Tucker Lewis Index, TLI), CFI, RMSEA, AIC, CAIC. (see

    Byrne, 1994, pp 53-58; Kline, Ch. 8, pp 189-214).

    12. Conventional model: Illustrating the input and output of the Children's Adjustment

    model (Model B: {PP10}). (cross-sectional model with multiple indicators). New

    elements: (a) How to model latent variables with several observed indicators for each

    latent factor, and (b), how to include demographics and other background variables as part

    of the model. (This model is described in the article by Pierce et al., 1998).

    We will also cover here the following additional topics:

    (a) how many indicators to model for each latent factor (see Marsh et al., 1998),

    (b) the rationale for indices prepared as parcels (groups) of items {PP11a, PP11b},

    (c) how to handle scales with different number of scale points {PP11c}.

    13. Basic rules of measurement and structural model identification {PP11d, PP11e, PP11f,

    PP11g, PP11h}. Reading: Kline, Chapter 6, pp. 125-149.

    14. Nested (hierarchical) and non-nested (non-hierarchical) models {PP12a, PP12b) and

    how to test the difference between nested models {PP12c, PP12d}. Reading: Kline,

  • 7/29/2019 Sem Helsinki2012

    5/18

    5

    Chapter 8, pp. 214-222.

    15. Reviewing the stages in constructing and testing a SEM model {PP13}. Reading: Kline,

    Chapter 5, pp. 91-95.

    16. Longitudinal panel model: Illustrating the input and output of the JOBS II mediationModel (Model D: {PP15a, PP15b, PP15c) New elements: (a) How you model effects of

    an experiment and its mediation effects {PP14c}, and (b), how to build a model that

    involves an intervention and longitudinal design of the follow-up assessments. (This

    model is described in the article by Vinokur and Schul, 1997).

    17. Estimating direct, indirect and total effects {PP16a, PP16b}. Reading: Kline, Chapter 7,

    pp. 164-169; and analyzing mediation effects {PP16b1 to PP16b5}.

    18. Modeling direct and mediating effects of an experiment using longitudinal design

    (Russell et al., 1998 {PP16c, d, e, and ee}.

    19. Group comparison model: Illustrating the input and output of the Job-SearchIntention Model (Model C; {PP14a, PP14b). New elements: (a) how to impose

    constraints to test equality of loading, exogenous variances and paths, and (b), how to

    compare the model across groups (e.g., is a model for males same or different than the

    model for the females). Imposing between group constraints (testing equality of factor

    structures, and of paths).

    20. How to model interactions (moderation effects)? {Wolchik et al., 1993 {PP16f} and

    Marsh et al., 2004).

    21. Multiple sources model: Illustrating the input and output of the Couples' Stress and

    Coping Model (Model E: {PP17a, PP17b, PP17c}). New elements: How to build a

    model with data from different sources (e.g., husbands and wives). Model E is the mostcomplex model with multiple indicators, longitudinal design and data from different

    sources. (This model is described in the article by Vinokur, Price and Caplan., 1996).

    22. Equivalent models and the issue of causal interpretation. {PP18a, PP18b, PP18c,

    PP18d}. (For a comprehensive article see MacCallum et al, 1993). Reading: Kline,

    Chapter 9, pp. 224-228.

    23. How to describe your model in a publication and what to report on?

    Based on two examples from the articles by Vinokur & Schul, (2002) , {PP19a, PP19b}

    and by Westman et al., (2005).{PP20a, PP20b, PP20c}

    24. How to model (and control for) common factor variance in cross-sectional design

    with measures from the same respondent. Based on the article by Vinokur-Kaplan,

    1995. {PP21a}, Also see Lindell & Whitney, 2001 and Williams & Anderson, 1994.

    25. Multi-trait Multi-method models and the correlated uniqueness model {PP21b,

    PP21c, PP21d (See Bagozzi, 1993; Kline, Chapter 9, pp 250-252). Also see Tomas &

    Oliver, 1999 {PP21e, PP21f}.

    26. Special Topics:

    (a) How to model change in longitudinal panel design {PP22a, PP22b}.

    Also an example from Price, Choi, Vinokur (2002) {P22c} also see Kessler and

    Greenberg (1981, pp 77-82), and Maassen & Bakker, 2001, pp 256 264.

  • 7/29/2019 Sem Helsinki2012

    6/18

    6

    (b) Confirmatory analysis: Do we need two factors or one would suffice?

    (E.g., Are social support and social undermining opposites of the same underlying

    latent factor? Is sense of mastery and optimism the same construct? (Marshal & Lang,

    1990{PP3, and PP12}). Or, for those who lose a loved one, is grief and depression the

    same psychological experience? (Vinokur, et al. (under review, 2002) {PP12cc}.

    (c) Confirmatory analysis: Is a given construct unidimensional? {PP24})(see Rubio et al, 2001).

    (d) Dealing with and interpreting correlations among errors and disturbances.

    (e) How to handle badly skewed variables, missing data, outliers, and categorical

    variables. (Kline, Chapter 3, pp 51-64, and Chapter 7, pp. 176-182).

    (f) How to deal with various error messages and nonconvergence. What to do when:

    a.... geting a warning message that certain parameters are linearly dependent on others,

    b.... when certain parameters were set to lowest or highest bound

    c.... you need to improve the fit of the model (dropping or adding parameters, retesting)

    (g) Suppressor variables in SEM models (see Maassen & Bakker, 2001); {PP22g,h, i,g}.

    (h) Power analysis for SEM models {PP23a}. Also see: Kline, Chapter 12, pp. 222-225.

    (i) Linear Growth Curve Modeling; {PP25a, PP25b, PP25c}

    (j) How to fool yourself with SEM? (Kline, Chapter 13, pp. 356-366) and {PP26}

  • 7/29/2019 Sem Helsinki2012

    7/18

    7

    II. Required, Highly Recommended Useful Readings1

    Required reading2 Highly Recommended readings, guidebooks or textbooks

    3 Very useful recommended readings or textbooks#

    available on the Ctools course website).

    A. Basic Textbooks and Manuals

    1Kline, R. B. (2011). Principles and practice of structural equation modeling. (3

    ndEdition). New

    York: The Guilford Press. ISBN for the paperback is: 9781606238769. [This is the main

    textbook for this course. Every student should have a copy. The local book stores

    will carry copies or else try Amazon.com].

    2Byrne, B. M. (2006, 2

    ndEdition). Structural equation modeling with EQS and EQS/Windows:

    Basic concepts, applications, and programming. Thousand Oaks, CA: Sage Publications.

    [Second most highly recommended textbook, but optional, for only if you intend to

    continue using EQS. Ideally, every student should have a copy or at least an access

    to the book in the library]

    Nearly all the content of the two manuals below is also embedded in the HELP section of

    the EQS software:2Bentler, P. M. (2006). EQS 6: Structural Equation Program Manual.

    Encino, CA: Multivariate software software [Content of this manual is also includedwith the EQS software under the Help menu]2Bentler, P.M., & Wu, E.J.C. (2002). EQS 6 for Windows User's Manual.

    Encino, CA: Multivariate software [Content of this manual is also included with the

    EQS software under the Help menu]

    B. Other useful textbooks

    Bollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.3Loehlin, J.C. (2004). Latent variable models: An introduction to factor, path, and structural

    equation analysis. (4

    th

    edition). NJ. Hillsdale: Lawrence Erlbaum Associates. [Textbook].3Schumacker, R.E., & Lomax, R.G. (2010). A beginner's guide to structural equation modeling.

    (3nd edition) N.J. Mahwah: Lawrence Erlbaum Associates. [Textbook]. In particular

    note Chapter 5 Goodness of fit criteria pp. 73-120.

    C. General reviews of SEM

    2 #Hoyle, R.H. & Smith, G.T. (1994). Formulating clinical research hypotheses as structural

    equation models: A Conceptual Overview. Journal of Consulting and Clinical Psychology,

    62, 429-440. [Among the best overview introductory articles to SEM].3, #

    Klem, L. (2000). Structural equation modeling. In L. G. Grimm, & P. R. Yarnold (Eds.),

    Reading and understanding more multivariate statistics . Washington, D.C.: American

  • 7/29/2019 Sem Helsinki2012

    8/18

    8

    Psychological Association.1, #

    Russell, D. W., Kahn, J. H., Altmaier, E.M., & Spoth, R. (1998). Analyzing data from

    experimental studies: A latent variable structural equation modeling approach. Journal of

    Consulting Psychology, 5(1), 18-29.2, # Vinokur, A.D. (2005). Structural Equation Modeling (SEM). In S. J. Best & B. Radcliff

    (Eds.), Polling America: An Encyclopedia of Public Opinion (pp. 800-805). Westport,

    Connecticut: Greenwood Press.

    D. Other Available Resources Dedicated to SEM

    Journals: Structural Equation Modeling: A multidisciplinary Journal Published by Lawrence

    Erlbaum Associates, Publishers.

    Websites:1. The website for Kline's Textbook: www.Guilford.com/Kline

    2. www.hawaii.edu/sem/sem.html

    3. http://davidakenny.net/kenny.htm (David Kenneys website).

    4 . SEMNET discussion group, see

    http://www2.gsu.edu/~mkteer/semnet.html

    To subscribe, see instructions on the above webpage

    4. To obtain data for your papers: http://www.icpsr.umich.edu See below:

    On a rich archive with thousands of documented datasets available to students and facultyfrom a consortium of about 400 universities around the world. Datasets can be used to prepare

    master theses, dissertations and other research publications.

    E. How to report SEM work in your research paper

    1,#Boomsma, A. (2000). Reporting analyses of covariance structures. Structural Equation

    Modeling, 7(3), 461-482.3, #

    McDonald, R.P., & Ho, M.R. (2002). Principles and practice in reporting structural equation

    analyses. Psychological Methods, 7 (1), 64-82.

    2, #Raykov, T., Tomer, A., & Nesselroade, J. R. (1991). Reporting structural equation modeling

    results in Psychology and Aging: some proposed guidelines. Psychology and Aging, 6(4),

    499-503. [Important article that set standards for how and what should be reportedin articles that include structural analyses but somewhat dated, Boomsma (2000)

    article is more up to-date and comprehensive]

    Example:3,#

    Vinokur, A.D., & Schul, Y. (2002). The web of coping resources and pathways to

    reemployment following a job loss. Journal of Occupational Health Psychology, 7, 68-83.

    F. Misc. Specific Topics

  • 7/29/2019 Sem Helsinki2012

    9/18

    9

    1. Issues relating to indicators, parcels and factors

    On Parceling:

    Bandalos, D.L. (2002). The effects of item parceling on Goodness-of-Fit and parameter estimate

    bias in structural equation modeling. Structural Equation Modeling, 9 (1), 78-102.

    Bandalos, D.L., & Finney, S.J. (2001). Item parceling issues in structural equation modeling. In.

    G.A. Marcoulides and R.E. Schumacker (Eds.) New developments and techniques in

    structural equation modeling. London: Lawrence Erlbaum Associates.

    Liang, J., Lawrence, R. H., Bennett, J. M., & Whitelaw, N. A. (1990). Appropriateness of

    composites in structural equation models. Journal of Gerontology: SOCIAL SCIENCES,

    45(2), 553-559.

    [The practical implications of this article for using one observed indicator (e.g.,

    index) for one latent variable are demonstrated in the article by Frone et al below.]

    Frone, M. R., Russell, M. and Cooper, M. L. (1992). `Antecedents and outcomes of work-family

    conflict: testing a model of the work-family interface', Journal of Applied Psychology,

    77(1), 65-78. Or, see Vinokur, Pierce and Buck (1999); Article placed in Ctools.

    Sass, D.A., & Smith, P.L. (2006). The effects of parceling unidimensional scales on structural

    parameter estimates in structural equation modeling. Structural Equation Modeling, 13

    (4), 566-586.

    Little, T.D., Lindenberger, U., & Nesselroade, J.R. (1999). On selecting indicators for

    multvariate measurement and modeling with latent variables: When good indicators are

    bad and bad indicators are good. Psychological Methods, 4 (2), 192-211.3,#Marsh, H.W., Hau, K., Balla, J.R., & Grayson, D. (1998). Is more ever too much? The number

    of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research,

    33(2), 181-220.

    Chen, F.F., S.G. West, and K.H. Sousa (2006). A comparison of bifactor and second-order

    models of quality of life. Multivariate Behavioral Research, 41(2): 189-225.

    Hall, R.J., A.F. Snell, and M.S. Foust (1999). Item parceling strategies in SEM: Investigating the

    subtle effects of unmodeled secondary constructs. Organizational Research Methods,

    2(3): 233-256.

    On Reflective and Formative Factors:

    Bollen, K.A., & Ting, K. (2000). A tetrad test for causal indicators. Psychological Methods, 5

    (1), 3-22.3.#Edwards, J.R., & Bagozzi, R.P. (2000). On the nature and direction of relationships between

    constructs and measures. Psychological Methods, 5(2), 155-174.

    Edwards, J.R. (2011). The fallacy of formative measurement. Organizational Research Methods,

    2011. 14(2): 370-388.

    Howell, R.D., E. Breivik, and J.B. Wilcox (2007). Reconsidering formative measurement.

    Psychological Methods, 12(2): 205-218.

    MacKenzie, S.B., Podaskoff, P.M., & Jarvis, C.B. (2005). The problem of measurement model

    misspecification in behavioral and organizational research and some recommended

    solutions. Journal of Applied Psychology, 90 (4), 710-730.

  • 7/29/2019 Sem Helsinki2012

    10/18

    10

    MacCallum, R.C., & Browne, M.W. (1993). Use of causal indicators in covariance

    structure models: Some practical issues. Psychological Bulletin. 114 (3), 533-541.

    Treiblmaier, H., P.M. Bentler, and P. Mair (2011). Formative constructs implemented via

    common factors. Structural Equation Modeling, 18(1): 1-17.

    Williams, L.J. and E. O'Boyle Jr. (2008). Measurement models for linking latent variables and

    indicators: A review of human resource management research using parcels. Human

    Resource Management Review, 18: 233-242.

    2. Adjustment for reliability1, #Bedeian, A. G., Day, D. V., & Kelloway, E. K. (1997). Correcting for measurement error

    attenuation in structural equation models: Some important reminders. Educational and

    psychological Measurement, 57 (5), 785-799.

    3. Model identification3, #Rigdon, E. E. (1995). A necessary and sufficient identification for structural modelsestimated in practice. Multivariate Behavioral Research, 30 (3), 359-383.

    4. Measures of goodness of fit and misfitHu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure

    analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6

    (4), 1-55.

    Browne, M. W., & Cudek, R. (1993). Alternative ways of assessing model fit. In K. A.

    Bollen, & L. S. Long (Eds.), Testing structural equation models. Newbury Park, CA:

    Sage.

    Costner, H.L. (1965). Criteria for measures of association. American Sociological Review,30 (3), 341-353.

    O'Boyle Jr, E.H. and L.J. Williams, (2011). Decomposing model fit: Measurement vs.

    theory in organizational research using latent variables. J. of Applied Psy., 96(1): 1.

    Williams, L.J. and E. O'Boyle Jr. (2011). The myth of global fit indices and alternatives for

    assessing latent variable relations. Organizational Research Methods, 14(2): 350-369.

    5. Model modification: capitalizing on chanceMacCallum, R. C., Roznowski, M., & Necowitz, L. B. (1992). Model modifications in

    covariance structure analysis: The problem of capitalization on chance. Psychological

    Bulletin, 111(3), 490- 504.

    6. Equivalent models3,#

    MacCallum, R. C., Wegener, D. T., Uchino, B. N., & Fabrigar, L. R. (1993). The problem of

    equivalent models in applications of covariance structure analysis. Psychological Bulletin,

    114(1), 184-199.

    7. Power analysis in SEM3,#MacCallum, R. C., Browne, M. W., & Sugawara, H.M. (1996). Power analysis and

    determination of sample size for covariance structure modeling. Psychological Methods, 1

    (2), 130- 149.

  • 7/29/2019 Sem Helsinki2012

    11/18

    11

    8. Modeling one or more common method effectsBagozzi, R. P. (1993). Assessing construct validity in personality research: Applications to

    measures of self-esteem. Journal of Research in Personality, 27, 49-87.Lindell, M.K., & Whitney, D.J. (2001). Accounting for common method variance in cross

    -sectional research designs. Journal of Applied Psychology, 86 (1), 114-121.

    [Demonstrating ways of dealing with negative affectivity and other common

    method effects also see below: Williams and Anderson]Podsakoff, P.M., Mackanzie, S.B., Lee, J., & Podsakoff, N.P. (2003). Common method

    biases in behavioral research: A critical review of the literature and recommended

    remedies. Journal of Applied Psychology, 88, 879-903.

    Vinokur-Kaplan, D. (1995). Treatment teams that work (and those that don't): An

    application of Hackman's group effectiveness model to interdisciplinary teams in

    psychiatric hospitals. Journal of Applied Behavioral Science, 31, 303-327.3,#

    Tomas J.M., & Oliver, A. (1999). Rosenberg's self esteem scale: Two factor or methodeffects. Structural Equation Modeling, 6(1), 84-98.

    Williams, L. J., & Anderson, S. E. (1994). An alternative approach to method effects by

    using latent-variable models: Applications in organizational behavior research. Journal

    of Applied Psychology, 79(3), 323-331. [Demonstrating ways of dealing with

    negative affectivity and other common method effects also see Lindell, M.K., &

    Whitney]

    9. Modeling data from different sources3,#Vinokur, A.D., Price, R.H., & Caplan, R.D. (1996). Hard times and hurtful partners: How

    financial strain affects depression and relationship satisfaction of unemployed persons and

    their spouses. Journal of Personality and Social Psychology, 71 (1), 166-179. [The

    Couples Stress and Coping Model: model E]. [Demonstrating analyses based on two

    sources of data as well as issues relating to longitudinal analyses, and comparison of

    alternative models, and models for different subgroups]

    10. Modeling Interactions (using categorical variables and

    continuous ones)Using categorical variables: Group comparisons

    1,#Benbenishty, R., Astor, R.A., Zeira, A, & Vinokur, A.D. (2002). Perception of violence

    and fear of school attendance among junior high school students in Israel. Social

    Work Research, 26(2) 71-87. [Article with comparisons across groups]

    Using continuous variables:

    Marsh, H.W., Wen, Z., & Hau, K. (2004). Structural equation models of latent

    interactions: Evaluation of alternative estimation strategies and indicator.

    Psychological Methods, 9 (3), 275-300.

    Jackman, M.G.-A., W.L. Leite, and D.J. Cochrane (2011). Estimating latent variable

    interactions with the unconstrained approach: A comparison of methods to form

    product indicators for large, unequal numbers of items. Structural Equation Modeling,

    18: 274-288.

    Wolchik, S. A., West, S. G., Westover, S., Sandler, I.N., Martin, A., Lustig, J., Tein, J., &

  • 7/29/2019 Sem Helsinki2012

    12/18

    12

    Fisher, J. (1993). The children of divorce parenting intervention: Outcome evaluation of

    an empirically based program. Amer. J of Community Psychology, 21 (3), 293-331.

    Aiken, L.S., & West, S.G. (1991), Multiple Regression: Testing and Interpreting

    Interactions. Newbury Park, CA: Sage Publications.

    11. Using Meta-analysis with SEMHome, P. W., Caranikas-Walker, F., Prussia, G. E., & Griffieth, R. W. (1992). A meta-

    analytic structural equation analysis of a model of employee turnover. Journal of

    Applied Psychology, 77 (6), 890-909. [Applying SEM to model the results of a

    Meta-Analysis]

    12. Modeling changeKessler, R.C., & Greenberg (1981). Linear panel analysis. New York: Academic Press.

    Chapter 6, pp. 77-80.

    Maassen, G.H., & Bakker, A.B. (2001). Suppressor variables in path models.Sociological Methods & Research, 30 (2), 241-270. [Regarding modeling change:

    see pp 256- 264.#Price, R. H., Choi, J., & Vinokur, A.D. (2002). Links in the chain of adversity following

    job loss: How economic hardship and loss of personal control lead to depression,

    impaired functioning and poor health. Journal Occupational Health Psychology, 7(4),

    302-312.#Russell, D. W., Kahn, J. H., Altmaier, E.M., & Spoth, R. (1998). Analyzing data from

    experimental studies: A latent variable structural equation modeling approach.

    Journal of Consulting Psychology, 5(1), 18-29. Addition to the article: SEE Slide

    PP16ee

    13. On causal inferences using SEMBollen, K. A. (1989). Structural equations with latent variables. New York: Wiley.

    Chapter 3, pp. 40-79.

    Pearl, J. (2000). Causality: Models, reasoning, and inference. Cambridge: Cambridge

    University Press.

    14. Mediational analysesBaron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable

    distinction in social psychological research: Conceptual, strategic and statistical

    considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

    Holmbeck G. N. (1997). Toward terminological, conceptual, and statistical clarity in the

    study of mediators and moderators: Examples from the child-clinical and pediatric

    psychology literatures. Journal of Consulting and Clinical Psychology, vol 65(4).

    Judd, C. M., & Kenny, D. A. (1981). Process analysis: Estimating mediation

    in treatment evaluations. Evaluation Review, 5, 602-619.

    MacKinon, D.P., Krull, J.L., & Lockwood, C.M. (2000). Equivalence of the mediation,

    confounding and suppression effect. Prevention Science, 1(4), 173-182.

  • 7/29/2019 Sem Helsinki2012

    13/18

    13

    Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social

    psychology. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), The handbook of social

    psychology (Vol. 1, 4th ed., pp. 233-265). Boston, MA: McGraw-Hill.

    Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental

    studies: New procedures and recommendations. Psychological Methods, 7(4), 422-

    445.

    Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects

    in structural equation models. In S. Leinhardt (Ed.), Sociological Methodology 1982

    (pp. 290-312). Washington DC: American Sociological Association.

    Preacher, K.J. and A.F. Hayes (2008). Asymptotic and resampling strategies for assessing and

    comparing indirect effects in multiple mediator models. Behavior research methods, 40(3):

    879-891.

    Kaplan Ed, R.M., Health Psychology, 2008. 27(2S): S99-S184. Special ISSUE on mediation.

    MacKinnon, D.P. (2008). Introduction to statistical mediation analysis. Erlbaum Psych Press.

    Also, more on Sobel test see: http://quantpsy.org/sobel/sobel.htm and for a test based on

    Bootstrap see: http://quantpsy.org/medmc/medmc.htm

    The following websites include abundance of information on MEDIATION and

    also on several topics of SEM:

    David Kenny's website (the first below) is most highly recommended.

    http://users.rcn.com/dakenny/kenny.htm

    http://www.gsu.edu/~mkteer/semfaq.html

    http://users.rcn.com/dakenny/mediate.htmhttp://www.public.asu.edu/~davidpm/ripl/mediate.htm

    15. Data preparation (transformations, outliers and imputations)Kline, R. B. (2011). Principles and practice of structural equation modeling. (3nd Edition).

    New York: The Guilford Press. Chapter 3, pp. 51-64, and Chapter. 7, pp. 176-182.

    16. Misc. modeling issues (e.g., setting weights, factor loadings)Gonzalez, R, & Griffin, D. (2001). Testing parameters in structural equation modeling:

    Every "one" matters. Psychological Methods, 6 (3), 258-269.

    Kaplan, D, & Ferguson, A.J. (1999). On the utilization of sample weights in latentvariable models. Structural Equation Modeling, 6(4), 305-321.

    Allison, P.D., (2003). Missing data techniques for structural equation modeling. Journal of

    Abnormal Psychology, 112(4): 545-557.

    Cole, D.A., J.A. Ciesla, and J.H. Steiger (2007).The insidious effects of failing to include

    design-driven correlated residuals in latent-variable covariance structure analysis.

    Psychological Methods, 12(4): 381-398.

  • 7/29/2019 Sem Helsinki2012

    14/18

    14

    G. Misc. References of Various Research Studies Using SEM

    Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of theAcademy of Marketing Science, 16(1), 74-94.

    Bryant, F. B., Edwards, J., Tin dale, R. S., Plosive, E. J., Heath, L., Henderson, E., &

    Suarez-Alcazar, Y. (Edited by). (1992). Methodological issues in applied social

    psychology. New York and London: Plenum Press.

    Dormann, C. (2001). Modeling unmeasured third variables in longitudinal studies.

    Structural Equation Modeling. 8 (4), 575-598.

    Farkas, A. J., & Tetrick, L. E. (1989). A three-wave longitudinal analysis of the causal ordering

    of satisfaction and commitment on turnover decisions. Journal of Applied Psychology, 74

    (855-868),

    Farrell, A. D. (1994). Structural equation modeling with longitudinal data: Strategies for

    examining group differences and reciprocal relationships. Journal of Counseling andClinical Psychology, 62(3), 477-487.

    Hays, R. D., Marshall, G. N., Wang, E. Y. I., & Sherbourne, C. D. (1994). Four-year

    cross-lagged associations between physical and mental health in the medical outcomes

    study. Journal of Consulting and Clinical Psychology, 3, 441-449.

    Krause, N., Liang, J., & Yatomi, N. (1989). Satisfaction with social support and depressive

    symptoms: A panel analysis. Psychology and Aging, 4(1), 88-97.

    MacCallum, R. C., & Browne, M. W. (1993). The use of causal indicators in covariance

    structure models: Some practical issues. Psychological Bulletin. 114(3), 533-541.

    Netemeyer, R. G., Johnston, M. W., & Burton, S. (1990). Analysis of role conflict and role

    ambiguity in a structural equations framework. Journal of Applied Psychology, 75,

    148-157.

    Newcomb, M. D. (1994). Drug use and intimate relationships among women and men: Separating

    specific from general effects in prospective data using structural equation models. Journal

    of Counseling and Clinical Psychology, 62(3), 463-476.

    Newsom, J.T., et al., (2008). Stable negative social exchanges and health. Health Psychology,

    27(1): 78-86.

    Pentz, M. A., & Chou, C. (1994). Measurement invariance in longitudinal clinical research

    assuming change from development and intervention. Journal of Counseling and Clinical

    Psychology, 62, 450-462.

    Rubio, D.M., Berg-Weger, M., & Tebb, S.S. (2001). Using structural equation modeling to

    test for multidimensionality. Structural Equation Modeling. 8 (4), 613-626.Schaubroeck, J., Cotton, J. L., & Jennings, K. R. (1989). Antecedents and consequences of role

    stress: A covariance structure analysis. Journal of Organizational Behavior, 10, 35- 58.

    Shaikh, A.R., et al., (2011). Direct and Mediated Effects of Two Theoretically Based

    Interventions to Increase Consumption of Fruits and Vegetables in the Healthy Body

    Healthy Spirit Trial. Health Education & Behavior, 38(5): 492-501.

    Wayne, S. J., & Ferris, G. R. (1990). Influence tactics, affect, and exchange quality in

    supervisor-subordinate interactions: A laboratory experiment and field study. Journal of

    Applied Psychology, 75, 487-499.#Westman, M., Vinokur, A.D., Hamilton, V.L., & Roziner, I. (2004). Crossover of Marital

    Dissatisfaction During Downsizing: A Study of Russian Army Officers and their Spouses.

    Journal of Applied Psychology, 89 (5), 769-779.

  • 7/29/2019 Sem Helsinki2012

    15/18

    15

    Williams, L. J., & Hazer, J. T. (1986). Antecedents and consequences of satisfaction and

    commitment in turnover methods: A reanalysis using latent variable structural equation

    methods. Journal of Applied Psychology, 71, 219-231.

    III. Articles for the course on Ctools Website(The articles on Ctools are under Course Resources subdirectory Course Readings)

    1. Bagozzi, R. P. (1993). Assessing construct validity in personality research: Applications to

    measures of self-esteem. Journal of Research in Personality, 27, 49-87.

    2. Bedeian, A. G., Day, D. V., & Kelloway, E. K. (1997). Correcting for measurement error

    attenuation in structural equation models: Some important reminders. Educational and

    Psychological Measurement, 57(5), 785-799.

    3. Benbenishty, R., Astor, R. A., Zeira, A., & Vinokur, A. D. (2002). Perceptions of violence and

    fear of school attendance among junior high school students in Israel. Social Work

    Research, 26(2), 71-87.

    4. Boomsma, A. (2000). Reporting analyses of covariance structures. Structural Equation

    Modeling, 7(3), 461-482.

    5. Chen, West, & Sousa (2006). A comparison of Bifactor and second-order models of quality of

    life. Multivariate Behavioral Research, 41(2), 189-225.

    6. Edwards, J. R., & Bagozzi, R. P. (2000). On the nature and direction of relationshipsbetween constructs and measures. Psychological Methods, 5(2), 155-174.

    7. Fincham, F. D., Beach, S. R. H., Harold, G. T., & Osborne, L. N. (1997). Marital

    satisfaction and depression: Different causal relationships for men and women?

    Psychological Science, 8(5), 351-357.

    8. Holmbeck G. N. (1997). Toward terminological, conceptual, and statistical clarity in the study

    of mediators and moderators: Examples from the child-clinical and pediatric psychology

    literatures. Journal of Consulting and Clinical Psychology, vol 65(4).

    9. Hoyle, R. H., & Smith, G. T. (1994). Formulating clinical research hypotheses asstructural equation models: A conceptual overview. Journal of Consulting and Clinical

    Psychology, 62(3), 429-440.

    10. Kessler, R.C., & Greenberg (1981). Linear panel analysis. New York: Academic Press.

    Chapter 6, pp. 77-80.

    11. Klem, L. (2000). Structural equation modeling. In L. G. Grimm, & P. R. Yarnold (Eds.),

    Reading and understanding more multivariate statistics. Washington, D.C.: American

    Psychological Association.

    12. Kline, R.B. (2005). How to fool yourself with SEM. In R.B. Kline: Principles and practice of

  • 7/29/2019 Sem Helsinki2012

    16/18

    16

    structural equation modeling. New York: The Guilford Press. (Ch. 12, pp 313-324).

    13. Maassen, G.H., & Bakker, A.B. (2001). Suppressor variables in models: Definitions and

    interpretations. Sociological Methods & Research, 30,(2), 241-270.

    14. MacCallum, R.C., & Browne, M.W. (1993). Use of causal indicators in covariance structure

    models: Some practical issues. Psychological Bulletin. 114 (3), 533-541.

    15. MacCallum, R. C., Browne, M. W., & Sugawara, H.M. (1996). Power analysis and

    determination of sample size for covariance structure modeling. Psychological Methods, 1

    (2), 130- 149.

    16. MacCallum, R. C., Wegener, D. T., Uchino, B. N., & Fabrigar, L. R. (1993).

    The problem of equivalent models in applications of covariance structure analysis.

    Psychological Bulletin, 114(1), 184-199.

    17. MacKenzie, S.B., Podsakoff, P.M., & Jarvis, C.B. (2005). The problem of measurement

    model misspecification in behavioral and organizational research and some

    recommended solutions. Journal of Applied Psychology, 90 (4), 710-730.

    18. Marsh, H.W., Wen, Z., & Hau, K. (2004). Structural equation models of latent interactions:

    Evaluation of alternative estimation strategies and indicator. Psychological Methods, 9

    (3), 275-300.

    19. Marshall, G. N., & Lang, E. L. (1990). Optimism, self-mastery, and symptoms of

    depression in women professionals. J. of Personality and Social Psych., 59(1), 132-139.

    20. McDonald, R. P., & Ringo Ho, Moon-Ho (2002). Principles and practice in reporting

    structural equation analyses. Psychological Methods, 7(1), 64-82.

    21. Noar, S. M. (2003). The role of structural equation modeling in scale development..

    Structural Equation Modeling, 10(4), 622-647.

    22. Pierce, P. F., Vinokur, A. D., & Buck, C. L. (1998). Effects of war-induced maternal

    separation on children's adjustment during the Gulf War and two years later. Journal

    of Applied Social Psychology, 28(14), 1286-1311.

    23. Podsakoff, P.M., Mackanzie, S.B., Lee, J., & Podsakoff, N.P. (2003). Common method

    biases in behavioral research: A critical review of the literature and recommended remedies.

    Journal of Applied Psychology, 88, 879-903.

    24. Price, R. H., Choi, J. N., Vinokur, A.D. (2002). Links in the chain of adversity following job

    loss: How financial strain and loss of personal control lead to depression, impaired

    functioning, and poor health. Journal of Occupational Health Psychology, 7(4), 302-312.

    25. Raykov, T., Tomer, A., & Nesselroade, J. R. (1991). Reporting structural equation

    modeling results in Psychology and Aging: Some proposed guidelines. Psychology and

    Aging, 6(4), 499-503.

  • 7/29/2019 Sem Helsinki2012

    17/18

    17

    26. Rigdon, E. E. (1995). A necessary and sufficient identification for structural models

    estimated in practice. Multivariate Behavioral Research, 30 (3), 359-383.

    27. Russell, D. W., Kahn, J. H., Altmaier, E.M., & Spoth, R. (1998). Analyzing data from

    experimental studies: A latent variable structural equation modeling approach. Journal of

    Consulting Psychology, 5(1), 18-29.

    28. Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:

    New procedures and recommendations. Psychological Methods, 7(4), 422-445.

    29. Tomarken, A. J., & Waller, N. G. (2003). Potential problems with well fitting models.

    Journal of Abnormal Psychology, 112 (4), 578-598.

    30. Tomas, Jose M., & Oliver, Amparo (1999). Rosenberg's Self-Esteem Scale: Two factorsor method effects. Structural Equation Modeling 6(1), 84-98.

    31. Vinokur, A. D., Pierce, P. F., & Buck, C. L. (1999). Work-Family Conflicts of Women in the

    Air Force: Their influence on Mental Health and Functioning. Journal of Organizational

    Behavior, 20, 865-878.

    32. Vinokur, A. D., Price, R. H., & Caplan, R. D. (1996). Hard times and hurtful partners:

    How financial strain affects depression and relationship satisfaction of unemployed

    persons and their spouses. J. of Personality and Social Psychology, 71 (1), 166-179.

    33. Vinokur, A. D., & Schul, Y. (1997). Mastery and inoculation against setbacks as activeingredients in the JOBS intervention for the unemployed. Journal of Consulting and

    Clinical Psychology, 65(5), 867-877.

    34. Vinokur, A. D., & Schul, Y. (2002). The web of coping resources and pathways to

    reemployment following a job loss. J. of Occupational Health Psych., 7(1), 68-83.

    35. Vinokur, A.D. (2005). Structural Equation Modeling (SEM). In S. J. Best & B. Radcliff

    (Eds.), Polling America: An Encyclopedia of Public Opinion (pp. 800-805). Westport,

    Connecticut: Greenwood Press.

    36. Vinokur-Kaplan, D. (1995). Treatment teams that work (and those that don't): Anapplication of Hackman's group effectiveness model to interdisciplinary teams in psychiatric

    hospitals. Journal of Applied Behavioral Science, 31, 303-327.

    37. Westman, M., Vinokur, A.D., Hamilton, V.L., & Roziner, I. (2004). Crossover of Marital

    Dissatisfaction During Downsizing: A Study of Russian Army Officers and their Spouses.

    Journal of Applied Psychology, 89 (5), 769-779.

    38. Vinokur, A.D., Merion, R.M., Couper, M.P., Jones, E.G., Dong, Y., (2006). Educational

    Web-Based Intervention for High School Students to Increase Knowledge and Promote

    Positive Attitudes Toward Organ Donation. Health Education and Beh., 33(6), 773-786.

  • 7/29/2019 Sem Helsinki2012

    18/18

    18

    39. Vinokur, A.D., Pierce, P. F., Lewandowski-Romps, L., Hobfoll, S., & Galea, S. (2011).

    Effects of war exposure on Air Force personnels mental health job burnout and other

    organizational related outcomes. Journal of Occupational Health Psychology, 16, 3-17.

    -------------------------------------------------- End of Syllabus ------------------------------------------