This working group is an informal gathering of qualitative researchers (students, faculty, and staff) at UNC.
The February session will focus on how tools in software to help with data visualization -- what Miles and Huberman call getting a "good view" of the data.
Registration is not necessary. If you have questions, please contact Paul_Mihas@unc.edu.
A similar group will meet at Duke's SSRI, tentatively on the second and fourth Thursdays of the month. Click here for more information or e-mail Alexandra Cooper at email@example.com.
The Duke group will meet in Duke's Gross Hall, SSRI West, room 230E. These meetings will begin on the 4th Thursday of Sept., 9/26, 12-2 p.m.
January 16, 2014 12:00 PM - 2:00 PM
February 20, 2014 12:00 PM - 2:00 PM
March 20, 2014 12:00 PM - 2:00 PM
Jennifer Wisdom and Kim Hoffman
The intent of this three-day workshop is to assist participants in the design of a rigorous mixed methods project. To this end, participants are encouraged to bring to the course a project to work on. This project may consist of (a) research study that they would like to complete using mixed methods; (b) a dissertation project that they see as mixed methods research; (c) a proposal for funding a mixed methods project; (d) an existing project that has been cast as a mixed methods study. Other project ideas will also be considered. Interspersed with the project development will be mini-presentations and discussion on the following topics: (a) an introduction to mixed methods research; (b) designing a mixed methods project; (c) writing a mixed methods journal article; (d) designing a proposal for funding (using NIH as a potential funding agency), and e) using qualitative software for analyzing data in a mixed me*thods project. The instructors will attempt to fit the class to the needs of the participants.
Fee: $1,300Registration details:
June 16, 2014 9:00 AM - 5:00 PM
June 17, 2014 9:00 AM - 5:00 PM
June 18, 2014 9:00 AM - 5:00 PM
This five-session short course provides an introduction to multilevel modeling for those who have no prior knowledge of the topic. It will also be useful for those with some experience in multilevel data analysis but a desire to develop a stronger theoretical grounding and improved understanding of different multilevel models and computer output. Multilevel modeling is used throughout the social, medical and other sciences to understand how response variables at one level of analysis can be influenced by variables from other levels in a data nested hierarchy. The R programming suite will be used to demonstrate the specification of various multilevel models. To register, click here. Instructor: Paul Voss.
To register, click here.
3010 Davis Library
March 05, 2014 2:00 PM - 4:00 PM
March 12, 2014 2:00 PM - 4:00 PM
March 19, 2014 2:00 PM - 4:00 PM
March 26, 2014 2:00 PM - 4:00 PM
April 02, 2014 2:00 PM - 4:00 PM
Structural equation models allow for the estimation of complex systems of equations and confirmatory factor analysis models. These models may include unobserved variables, which may involve conditions that violate the assumptions of ordinary regression, including imperfect measurement, correlated error amongst indicator variables and between equations, and reciprocal causation. This course will present a series of applied examples using EQS syntax under CALIS, which allows a complex structural equation system to be represented by a series of straightforward, regression-like equations. For more information, contact Cathy Zimmer, 962-0516.
To register, click here.
This short course provides an introduction to logistic regression. Model specification, identification, estimation, hypothesis-testing, and interpretation of results are covered. We meet in 14 Manning Hall. Software to estimate these models is discussed, but not demonstrated. This is not a course on software, but rather a course on the concepts and uses of logistic regression.To register, click here.
For course handouts, click here.
March 19, 2014 3:00 PM - 5:00 PM
A powerful method for analyzing longitudinal data is Latent Trajectory Analysis (LTA). LTA allows each case in a sample to have individual trajectories ("latent curves" or "growth curves") representing change over time. In addition to mapping these trajectories, LTA allows researchers to examine the determinants of these trajectories or to relate the trajectories of one variable with those of another. The approach to LTA in this course draws on the strengths of structural equation modeling (SEM), and the primary goal is to introduce participants to the theory and application of LTA. The course begins with a conceptual introduction to LTA, a description of research questions that are well-suited for the technique, and a review of SEMs. The remainder of the course will cover the following topics: LTA models for a single variable with and without predictors of differences in trajectories; modeling nonlinear trajectories; the LTA model for multiple variables; the relation between the parameters governing the trajectories in two or more variables; incorporating predictors of multiple trajectories; and extensions to the LTA model. Participants should have prior training and experience with structural equation modeling and related software.
Instructor: Kenneth Bollen.
Class is from 9 a.m. to 5 p.m. For registration details, click here
June 02, 2014 9:00 AM - 5:00 PM
June 03, 2014 9:00 AM - 5:00 PM
June 04, 2014 9:00 AM - 5:00 PM
June 05, 2014 9:00 AM - 5:00 PM
June 06, 2014 9:00 AM - 5:00 PM
The Growth Mixture Model (GMM) is an extension of the Latent Growth Curve Model (LGCM) that identifies distinct subgroups of growth trajectories and allows individuals to vary around subgroup-specific mean trajectories. Conventional growth modeling estimates a single mean intercept and slope for each individual and variance parameters around the mean intercept and slope. The GMM relaxes the assumption that all individuals are drawn from a single population with common parameters by using latent trajectory classes, resulting in separate intercepts, slopes, and variance parameters for each subgroup.
This workshop will provide training in estimating GMMs to analyze growth trajectories. Key features of this model are that it can identify the number and form of distinct subgroups of growth trajectories, estimate the proportion of the population in each subgroup, and model predictors of the trajectories and predictors of class membership. In addition to the basic model, this workshop will cover several extensions, such as including a distal outcome predicted by the trajectories, multiple group GMMs , and parallel process or joint trajectory models.For registration details, click here
June 09, 2014 9:00 AM - 5:00 PM
June 10, 2014 9:00 AM - 5:00 PM
June 11, 2014 9:00 AM - 5:00 PM
Prior training in maximum likelihood estimation. At least introductory knowledge of R preferred.
The course will be an introduction to Bayesian statistics. We will meet for three full days, with morning and afternoon sessions of 2.5-3 hours. We will focus on the two complementary goals of learning the theory behind Bayesian inference as well as practical implementation in R. Students will walk away from the course with an understanding of how to apply Bayesian models and what is going on “under the hood” with their results. Class time will be spent in lecture and working hands-on with example data in R.
Upon completion of the course, students will have enough knowledge to use Bayesian methods in their own research. This new skillset will broaden the scope of research questions they can seek to answer, expand their ability to conduct inference to answer those questions, and make presentations of their research (such as in job talks) more compelling to hiring departments.
Morning: Introduction; Probability Review; The Logic of Bayesian Inference
Afternoon: Prior Information and Prior Distributions
Morning: Model Checking and Diagnostics; Model Comparison Afternoon: Markov-chain Monte Carlo (MCMC) Theory; Simulation-Based Inference
Morning: Hierarchical Bayesian Models
Afternoon: Estimating, Summarizing, and Presenting Bayesian Models in R
Registration details TBA.
July 28, 2014 9:00 AM - July 30, 2014 5:00 PM
This one-day short course will provide participants with an introductory overview of issues frequently encountered when conducting secondary computer analyses of data collected from sample surveys with complex multi-stage designs (e.g., PSID, NHANES, NCS), including design-based weight determination, software choice, and proper analysis methods. The workshop is not intended for participants looking to design a survey, but rather for participants who have a desire to analyze complex sample survey data. TOPICS:
- Recognizing a sample with a complex design, and sampling error calculation models
- Calculation of sample weights based on sample designs / non-response / post-stratification
- Calculation of new weights for subgroups / longitudinal analyses
- Weighted vs. unweighted analyses
- Variance estimation, and calculation of correct confidence intervals for population parameters
- Hypothesis Testing based on sample estimates
- Design Effects
- Software packages capable of complex sample survey data analysis
- Common analysis methods (linear modeling, descriptive statistics), interpretation of results
- Using weights in regression modeling, and model-based approaches
- Examples using software programs to analyze real survey data
CPSM Students: $30
UNC Students: $45
This course will count as 7.0 CPSM short course credit hours.
Emily McFarlane Geisen and Amanda Wilmot
Focus group interviews are commonly used for survey development, content development, and qualitative data collection to capture rich information about attitudes and beliefs that affect behavior. An overview of the basics of focus groups supplemented with real examples and hands-on practice will highlight the most appropriate uses of focus groups, moderating focus groups, developing interview questions, analyzing and using results, as well as reporting findings.
To register, click here
CPSM students: $30
UNC students: $45
This course will count as 7.0 CPSM short course credit hours
Mplus is a modeling program that integrates random effect, factor, SEM and latent class analysis in both cross-sectional and longitudinal settings and for both single-level and multi-level data. As such, this short course will only scratch the surface of Mplus' capabilities. The basic structure of the program and how it can be modified will be taught in a hands-on way in the Odum Institute Computer Lab. For a handout, click here.
March 26, 2014 3:00 PM - 5:00 PM
Although graduate school teaches you many things, many Ph.D. students complete their dissertations having never been fully trained in academic publishing. How does academic publishing actually work? Where should you publish your research? How do you actually get papers accepted in strong academic journals? In this talk, we will discuss several major ideas in academic publishing (focusing on the social sciences). We will discuss:
- professional integrity and ethics
- the role of academic conversations and communities as a guide for publication forums and journal selections
- the mechanics of publishing in journals and other forums (including outlining, writing style, journal, legal, and newspaper submissions, the peer review process, revisions, and corresponding with editors)
- and acceptance and all that follows.
UNC students: $15
UNC staff/faculty: $30
To register, click here.
April 04, 2014 1:30 PM - 4:00 PM
This course offers an introduction to a new analytical method of agent-based modeling. It provides an easy way for the beginner to translate research goals into a dynamic model in simulation form. This course offers a step-by step, hands-on, interactive approach to conceptualizing, creating, implementing, and analyzing policy/political science simulation models. This analytical tool can be used in addition to traditional triangulation strategies to operationalize quantitative and qualitative variables (or a combination of both) into a simulation.$30, UNC Students
$40, UNC Faculty/Staff
Registration details TBA.
April 30, 2014 10:00 AM - 4:00 PM
Network Analysis: Statistical Inference with Exponential Random Graph Models Exponential random graph models (ERGMs) are flexible statistical models for relational (i.e., network) data that are capable of representing and identifying an extensive range of interdependencies common in networks. Are gender, race or social class predictive of mixing patterns in a social network? Are ties in a directed network typically reciprocated? Does the network exhibit transitive triad closure? The simultaneous and stochastic manifestation of relational dependencies such as these can be identified and characterized with ERGMs. This workshop will introduce ERGM and demonstrate its application in the free and open source R statistical software. Participants will be provided with real-world network data as well as R code to apply ERGMs to that data.
Prerequisites for this course include a background in basic network analysis concepts and novice experience with the R statistical software.
Registration will open 60 days before the course. Fees TBA.
If you have questions, please contact Paul_Mihas@unc.edu.
3010 Davis Library
May 08, 2014 10:00 AM - 4:00 PM
Doug Steinley, Missouri
Network analysis focuses on relationships between or among social entities. It is used widely in the social and behavioral sciences, as well as in political science, economics, organizational studies, behavioral biology, and industrial engineering. The social network perspective, which will be taught in this workshop, has been developed over the last sixty years by researchers in psychology, sociology, and anthropology. The social network paradigm is gaining recognition in the social and behavioral sciences as the theoretical basis for examining social structures. This basis has been clearly defined and the paradigm convincingly applied to important substantive problems. However, the paradigm requires concepts and analytic tools beyond those provided by standard quantitative (particularly, statistical) methods. This five day workshop covers those concepts and tools. The course will present an introduction to concepts, methods, and applications of social network analysis drawn from the social and behavioral sciences. The primary focus of these methods is the analysis of relational data measured on groups of social actors. Topics include an introduction to graph theory and the use of directed graphs to study actor interrelations; structural and locational properties of actors, such as centrality, prestige, and prominence; subgroups and cliques; equivalence of actors, including structural equivalence, blockmodels, and an introduction to relational algebras; an introduction to local analyses, including dyadic and triadic analyses; and an introduction to statistical analyses, using models such as p1 and exponential random graph models. The workshop will use several common software packages for network analysis: UCINET, Pajek, NetDraw, and STOCNET.
The course runs from 9 a.m. to 5 p.m. Monday-Friday with a break for lunch.
For registration details, click here.
August 04, 2014 8:30 AM - 5:00 PM
August 05, 2014 9:00 AM - 5:00 PM
August 06, 2014 9:00 AM - 5:00 PM
August 07, 2014 9:00 AM - 5:00 PM
August 08, 2014 9:00 AM - 5:00 PM