College of Arts & Sciences-Department of Sociology

UNIVERSITY OF NORTH TEXAS

COLLEGE OF ARTS & SCIENCES-DEPARTMENT OF SOCIOLOGY

SOCI 6200: SEMINAR ON RESEARCH METHODS-ADVANCED STATISTICS

Fall 2016

Instructor: Gül Seçkin, Ph.D. Class Dates: Thursday

Office: Chilton Hall 390D Class Time: 6:00p.m-8:50 p.m.

Email: Class Location: Chilton 274

Phone: 940-369-7751

Office Hours: Thursday 3:30-5:30 p.m.

COURSE OVERVIEW

The focus of this course is applied statistical techniques, with emphasis on linking measurement and structural models through structural equations with latent constructs. I have chosen the content of the course to redress what I perceive as an unfortunate disparity between students' exposure to statistics discussed in the abstract vs. statistics applied first hand to real-life data sets. In my own (and others') experience, real understanding of statistical procedures does not start until one actually applies the statistical techniques to concrete data. Consequently, in this course you will complete weekly computer assignments that provide “hands-on” experience with data analysis.A secondary focus of the course is to enhance students' ability to interpret and critically evaluate the methodology used in journal articles, particularly data analytic techniques. The ability to scrutinize others' work increases students' skills in applying appropriate procedures to their own research. A further benefit of critiquing research reports is the increased opportunity for publications, either through a redesign and replication of the suspect study, or a secondary analysis of the existing data set. Hopefully, this course will enhance your ability to mesh substantive thinking with appropriate statistical procedures.

COURSE BOOK

Mertler, C. &Vannatta, R (2010). Advanced and Multivariate Statistical Methods. Pyrczak Publishing.

Additional Weekly readings will be provided

COURSE REQUIREMENTS

Weekly Computer Exercises:I will provide a list of weekly computer assignments in which you interpret output from your own data analyses. There will be six weekly assignments, 20 pointseach. These homeworkswill be designed for you to demonstrate conceptually what you have learned each week. Additionally, they will ask you to exercise your skills in application, analysis, synthesis, and critical evaluation of research data.Prepare figures and tables to present your findings. You will also turn in your computer output with your homework.Include all your data analysis output, properly paginated so I can immediately find the section of output you’re using to support the results reported in your paper.

Take-Home Midterm:worth 100 pointsof your grade. The purpose of this assignment is to provide practice in interpreting and evaluating the data-analytic procedures of published research. In midterm, you will critically evaluate the methods of data analysis used in a published study. The data analysis should ideally include structural equation modeling (SEM), possibly supplemented by other techniques such as multiple regression or exploratory factor analysis. Choose a study published in mainstream journals or books. Your exam should not be longer than 12double-spaced, typewritten pages (using a 12 pt. font. You will reanalyze in your final, the data set of the published study you selected for your midterm. Therefore, you must make sure that the data set is available to you. For this reason, you can pick a study to critique that analyzes at least one year of the General Social Survey (GSS). Taken almost every year since 1972, the GSS is a nationally-representative data set of U.S. adults. It contains numerous variables related to attitudes, behavior, and personal characteristics. You are of course free to critique a published paper for your midterm that do not use the GSS if you already have the data set in your possession, or you can guarantee me you will have it within a few weeks of the semester.Further written instructions about your midterm will be posted on course blackboard site. You will have one week to complete your exam.

Take-Home Final: worth 100 points of your grade. Your final requires replication of a research study with data analysis(e.g., analysis of a data set to test new hypotheses. You should begin by attempting to replicate one or more SEM analyses in the original study. If the original study used a regression analysis only, I want you to extend that analysis by conducting a CFA and/or SEM.You will describe the findings from your efforts to replicate and extend the data analyses. Prepare figures and tables to present your findings. You will also turn in your computer output with your final. Include all your data analysis output, properly paginated so I can find the section of output you’re using to support the results reported in your paper. The exam should be no more than 15 typewritten, double-spaced pages (excluding computer output and tables of results).Further written instructions about your final will be posted on course blackboard. You will have one week to complete your exam.

Presentation:80 points of your grade. You will present a power presentation of your final paper(e.g., introduction, sample, research design, measurement and structural model, data analyses (SEM and other analytical models), results, and interpretation of your findings). In describing these components, you will note both strengths and shortcomings of the original study, and what aspects of the investigation you tried to improve.

WEEKLY SCHEDULE OF READINGS & ASSIGNMENTS

September 1: ASSUMPTIONS IN MULTIVARIATE ANALYSIS

Read Chapter 3 in Mertler