UNC, PSYC 236

Excerpt: ... Psychology 236: Factor Analysis Course Outline Fall, 2004 Instructor: Robert MacCallum 353B Davie Hall 962-4016 Email: maccallum@unc.edu Office hours: Tuesdays 9:00-10:00 or by appointment Wednesdays, 9:00-11:45; Peabody 306 Class Meetings: Course Materials: Packet of course notes available at UNC Student Stores. Software: (1) Exploratory factor analysis software CEFA; Download from http:/quantrm2.psy.ohio-state/browne/ (2) Confirmatory factor analysis software LISREL or RAMONA; Access to be provided. Additional material will be provided online at http:/www.unc.edu/~rcm/. Grading: Homework 30%; Midterm Exam 35%; Final Exam 35% Schedule of Topics: Meeting 1 2 3 4 5 6 7 8 9 10 Date 8/25/04 9/1/04 9/8/04 9/15/04 9/22/04 9/29/04 10/6/04 10/13/04 10/20/04 10/27/04 Topics Orientation and Introduction History and Models Matrix Algebra The Common Factor Model Fitting the Common Factor Model to Data CEFA Software Estimating the Number of Common Factors The Rotation Problem: Orthogonal Analytical Rotation Midterm ...

Maryville MO, NYUTUP 030509

Excerpt: ... THE DEVELOPMENT OF THE STUDENT COUNSELING NEEDS SCALE Pius N. Nyutu Dr. Norman C. Gysbers, Dissertation Supervisor ABSTRACT The purpose of this study was to develop an instrument that can be used in the identification of guidance and counseling needs of secondary school students in Kenya. The Student Counseling Needs Scale (SCNS) is a 52-item inventory. Data from 867 participants (423 males and 444 females) recruited from seven provincial schools in Kenya was analyzed using confirmatory factor analysis (CFA) and exploratory factor analysis (EFA). Five factors, human relationships, career development, self development, social values, and learning skills were assessed. Alpha coefficients for the five SCNS subscales ranged from .83 to .88, and .94 for the whole scale. Additional analysis revealed differences by gender and school in the way students rate their guidance and counseling needs. The findings highlighted the importance of using assessment instruments in identification of students' counseling needs. Th ...

Minnesota, PSY 8815

Excerpt: ... Week 6: Confirmatory Factor Analysis I. Confirmatory Factor Analysis A. Testing a model that was specified in advance (either with a strong theoretical or empirical basis) B. Use a path model to represent the hypothesized factor structure 1. Rectangles = observed variables 2. Circles = latent variables 3. Arrows = paths C. Matrices represent the elements lambdas, deltas, etc. 1. matrix has as many rows as observeds, and as many columns as latents 2. is a symmetric square matrix 3. is also a symmetric square matrix II. LISREL and examples (from LISREL guides) A. LISREL steps 1. Create syntax and modify: Setup Build LISREL syntax a) Enter in data matrix b) Add options if necessary 2. Run 3. Interpret output B. Example 1: Holzinger & Swineford (1939) collected data on twenty-six psychological tests administered to 145 seventh- and eighth-grade children in the Grant-White school in Chicago. Six of these tests were selected and for this example it was hypothesized that these measure two common facto ...

Wisconsin Milwaukee, EDPSY 825

Excerpt: ... ED PSY 825 / Razia Azen 1 A brief introduction to confirmatory factor analysis The main difference between exploratory and confirmatory factor analysis is that exploratory factor analysis (EFA) is data driven whereas confirmatory factor analysis (CFA) is theory driven. This means that in CFA the researcher must specify how the variables should load on the factors prior to the analysis and then use the analysis to test this theory/model. o The theory/model can be modified based on the data, but too many data-based modifications make this an exploratory procedure! o Results should be cross-validated, preferably with another sample. CFA and SEM - Confirmatory Factor Analysis is included within the larger and more general Structural Equation Modeling (SEM) framework. - SEM involves a "measurement model" component and a "structural model" component. - The measurement component allows the researcher to model the error with which observed variables are measured and relate the observed variables to underlying const ...

Idaho, PSYC 430

Excerpt: ... Psychology 430 Construct Validity Construct Validity Define the construct Establish a nomological network ".to establish its relation to other variables with which it should, theoretically, be associated positively, negatively, or practically not at all" (Westen & Rosenthal, 2003) Convergent and discriminant validity Demonstrating Construct Validity Changes occur as predicted "K "Known groups" study " d Time or age Experimental manipulations (also known as "method of contrasted groups") Multitrait-multimethod design (MTMM) Multiple traits or other characteristics are measured by multiple methods Convergent & discriminant validity evidence 1 Demonstrating Construct Validity (cont.) Factor analysis Exploratory factor analysis (EFA) p y y Confirmatory factor analysis (CFA) You specify the factors Can test the fit of your proposed model Factor Analytic Model of Variance Total variance = Common variance + specific variance + error variance Common variance is called the "communal ...

Portland, PSY 624

Excerpt: ... Psychology 524/624 Lab Assignment # 3 Due in Class before Midterm Exam: April 25, 2008 Confirmatory Factor Analysis 1. Conduct a confirmatory factor analysis to test whether and how well a two factor solution fits the humor data. Use the data set humor.sav. 1a. Report and interpret the X2. 1b. Report and interpret the GFI. 1c. Report each of the path coefficients and whether or not they are significant. 1d. Compare your results to the single factor Allen and single factor Rickles analyses. What do you notice? What would you conclude and why? Validity 2. Frank has created two scales assessing Rickles and Allen type humor. He has also discovered two scales that assess sadism and self-deprecation. Show Frank how he can use these scales to assess the validity of the Rickles and Allen scales. Include and interpret the results of your validity analyses. Use the data set newhumordata.sav. Measure Development Example 3. Find an article that reports on the development of a measurement instrument that uses ...

Maryland, EDMS 722

Excerpt: ... _ You are expected to get data from some source, presumably a source in your field of interest. These data may be from a study that you, your colleagues, or your advisor are working on or have worked on. The data may be from a public database, or from a published study in which descriptive information (i.e., correlations and standard deviations, and means if necessary) are given. Acceptable types of projects Fairly elaborate path analysis model. Confirmatory factor analysis , usually for the purpose of scale validation. Single factor models are not acceptable. Testing competing models as part of the project is desired. Structural models with latent factors (should have three or more factors). Multi-group path analysis models - if you have data from two or more groups and would like to test if the path models are identical. Multi-group confirmatory factor models - if you have data from two or more groups an ...

NYU, G 89

Excerpt: ... Structural Equation Methods G89.2247 Due 04/05/04 Exercise 8 The data set SEM8.txt contains seventeen variables derived from Okazaki's study. The first is ethnic group (0=Anglo, 1=Asian-American), the second is sex (which will not be used in the analysis), and the remaining fifteen variables are sets of three indicators for independence, interdependence, depression (BDI), social avoidance (SADS) and fear of negative evaluation (FNE). I constructed the sets of three indicators by carefully balancing the content of the items so that each indicator is as close as I could get to a "parallel form" or optimal "split third". 1. Using the SEM software of your choice, carry out a confirmatory factor analysis of the five sets of three indicators of independence, interdependence, BDI, SADS and FNE. Include the ethnic variable in this analysis so that you can estimate the correlation of ethnic with the five latent variables. A. Comment on the fit of this CFA model. B. Compare the correlations of the ethnic indicator an ...

North Texas, RSS 6810

Excerpt: ... ables'. In Multiple Regression and Beyond, Keith, 2006 Supplemental Readings Proactive Coping, Positive Affect, and Well-Being, Greenglass and Lisa Fiksenbaum, 2009 Coping Style Use Predicts Posttraumatic Stress and Complicated Grief Symptom Severity Among College Students Reporting a Traumatic Loss, Schnider, Elhai, and Gray Notes Path Analysis Recommended Texts Principals and Practice of Structural Equation Modeling, Kline Multiple Regression in Behavioral Research, Pedhazur Multiple Regression and Beyond, Keith Chapter 13 Factor Analysis Readings Factor Analysis, Darlington Supplemental Readings The Authentic Personality, Wood, Linley, Maltby, Baliousis, Joseph Notes Factor Analysis Recommended Texts Analyzing Multivariate Data, Lattin, Carroll, and Green Introduction to Factor Analysis, Kim and Mueller Factor Analysis, Kim and Mueller Factor Analysis and Related Methods, McDonald Chapter 14 Advanced Techniques: Confirmatory Factor Analysis , Structural Equation Modeling, Multilevel Modeling Supplemental ...

CSU Northridge, PSY 524

Excerpt: ... Structural Equation Modeling Intro to SEM Psy 524 Ainsworth AKA SEM Structural Equation Modeling CSA Covariance Structure Analysis Causal Models Simultaneous Equations Path Analysis Confirmatory Factor Analysis SEM in a nutshell Combination of factor analysis and regression Continuous and discrete predictors and outcomes Relationships among measured or latent variables Direct link between Path Diagrams and equations and fit statistics Models contain both measurement and path models Jargon Measured variable Observed variables, indicators or manifest variables in an SEM design Predictors and outcomes in path analysis Squares in the diagram Latent Variable Un-observable variable in the model, factor, construct Construct driving measured variables in the measurement model Circles in the diagram Jargon Error or E Variance left over after prediction of a measured variable Variance left over after prediction of a factor Variable that predicts other variables Disturbance or D ...