Beginning in fall of 2010, the CRMDA has offered Saturday afternoon software Seminars. These seminars are typically 3-hour workshops designed to familiarize researchers (including students and faculty) or other interested people with the basics of particular software packages (e.g., Mplus, LISREL, R, Stata, SAS) or methodological procedures (e.g., Multiple Imputation). They take place in Watson Library, Room 455 (unless otherwise noted). Seminars are free to anyone who would like to attend. Seminars are run by 2-3 graduate students who are high-end users of the software. Participants are welcome to come simply to learn the basics, or with particular questions in mind.
We encourage you to join our announcement list serve (Methods-L) so you can be informed of events and talks related to methodology. Specifically, Methods-L is an announcement list whereby we periodically inform the KU community about methodology related events (such as the Saturday Seminar Series) or activities that may be of broad interest to researchers at KU. To join the Methods-L list serve please send your name and email address to the CRMDA
CRMDA Saturday Seminar Schedule (Spring 2013) 1:00 - 4:00 PM in Watson 455 (unless otherwise noted)
February 9th
Introduction to R, by Sunthud Pornprasertmanit, Terry Jorgensen, & Fan Jia
10 Budig Hall (Budig Instruction Lab)
This seminar will introduce the R statistical computing environment. R is an open-source (free!), programmable statistics platform with many options for different types of data analysis. The topics for this seminar include: (1) Importing and exporting data, (2) Getting around in R (the R console, objects, basic computation), (3) Analyzing data, (4) Visualizing data, (5) Programming in R, (6) Extending your R knowledge (packages, help) and (7) A look ahead to more advanced topics (linear regression, 3-d graphics). These topics will be addressed through hands-on activities with example data sets and R code templates that you can take and re-use for your future projects. Additionally, time will be available during breaks and for a short period after the seminar for individual consultations.
Contact us (dspainhower@ku.edu) by February 7th if interested
February 16th
Regression and ANOVA Analyses in R, by Sunthud Pornprasertmanit, Terry Jorgensen, & Fan Jia
This seminar will extend the knowledge from the Introduction to R class. The topics for this seminar include: (1) simple and multiple regression, (2) interactions and categorical variables multiple regression, (3) diagnostic checks for violation of assumptions, (4) one-way and factorial analysis of variance, and (5) visualizing data for further interpretation. Additional topics (e.g., mediation) may be included as time allows. The class is designed such that any persons who have experienced R for about two hours can understand the course materials. These topics will be addressed through hands-on activities with example data sets and R code templates that you can take and re-use for your future projects. Additionally, time will be available during breaks and at the end of the seminar for individual consultations.
Contact us (dspainhower@ku.edu) by February 14th if interested
March 9th
Introduction to Missing Data, by Kyle Lang
This seminar will cover the fundamental principles of modern missing data analysis with special attention given to the conditions needed to correctly implement the current best practice approaches. Discussion will focus on the benefits of modern principled techniques over antiquated ad hoc approaches, and the practical details of the estimation procedures underlying important missing data treatments.
The topics covered will include a brief introduction to the missing data problem including an overview of patterns of missingness and missing data mechanisms. Ad hoc approaches to missing data analysis will also be discussed along with the potential biases that can arise from their implementation. Finally, the most appropriate, modern missing data techniques will be introduced with a special focus on a stepwise explanation of their operation. In this last portion of the talk, the Expectation Maximization (EM) Algorithm, Multiple Imputation (MI), and Full Information Maximum Likelihood (FIML) will be covered. If time allows, we will cover syntax examples to demonstrate the implementation of these three techniques.
By the end of the session, all attendees should leave with a general understanding of the process by which the various missing data algorithms operate when implemented in substantive analyses.
Contact us (dspainhower@ku.edu) by March 7th if interested
April 13th
REDCap, by Ada Van Roekel and Skyler Adams
Seminar will be held in Watson 419
This seminar will help researchers learn how to construct a REDCap database and survey in order to collect, manage, and analyze their data in the most secure and organized way possible. While there are other database and survey tools available, REDCap is certified HIPAA compliant and provides a user-friendly interface for building both surveys and databases. The topics for this seminar include: 1) An introduction and overview of REDCap and REDCap survey, 2) Building various types of REDCap databases and surveys, 3) Importing and exporting data from a survey or database, 4) Managing user rights, 5) Using the scheduling module, 6) How to use the forms library to save time, and 7) Building reports. These topics will be addressed through hands-on activities with example data sets. The goal of this seminar is to introduce students, faculty, and staff to an efficient, effective way to capture, manage, and analyze their data. Time will be available for a short period after the seminar for individual consultations.
Contact us (dspainhower@ku.edu) by April 11th if interested
April 20th
LISREL, by Luke McCune & Beth Grandfield
This seminar will demonstrate how to estimate a confirmatory factor analysis (CFA) model using the LISREL software package. While LISREL may seem more complicated than other CFA packages, the matrix approach it takes is easy to master with a little practice and some great examples. The topics for this seminar include: (1) An introduction to LISREL, (2) Importing data, (3) creating syntax, (4) interpreting output, and (5) some common trouble shooting tips, including starting values and common error messages. These topics will be addressed through hands-on activities with example data sets; though you are encouraged to bring your own data set to work on. The goal of this seminar is to introduce students, faculty and staff to an effective way to estimate SEM models using LISREL. We will provide access to reference guides for importing data, creating syntax, interpreting output, etc. Additionally, time will be available during breaks and for a short period after the seminar for individual consultations.
Contact us (quant@ku.edu) by April 18th if interested
April 27th
Mplus, by Aaron Boulton & Steve Short
Mplus (Version 6.0, Muthén & Muthén, 2010) is a powerful statistical modeling program that provides researchers from all disciplines with an array of tools for general data analysis. The program draws from a general modeling framework that allows for the evaluation of complex models using both observed and unobserved (i.e. latent) variables. The researcher is offered a wide variety of models, estimators, and algorithms to facilitate use, and Mplus readily accommodates differing variable types, missing data, and complex sampling designs. Furthermore, Mplus features a straight-forward programming language that most researchers find quite approachable. The purpose of this seminar is to provide an introduction to Mplus for beginners while offering experienced users an additional perspective on program use and functionality. The seminar will proceed as follows: First, an overview of the Mplus programming language and data importation functions will be discussed. Next, a variety of statistical procedures, including but not limited to multiple regression, path analysis, exploratory/confirmatory factor analysis (EFA and CFA), structural equation modeling (SEM), multilevel modeling, longitudinal modeling (growth curves), mixture modeling (latent class analysis) will be discussed and demonstrated with examples. The seminar will end with a group Q&A session regarding the material and individual project troubleshooting. Attendees are assumed to have some familiarity with basic statistics including regression analysis. Knowledge of latent variable analysis (structural equation modeling) will enhance understanding but is not required.
Contact us (quant@ku.edu) by April 25th if interested
May 4th
Advanced Stata, by Jacob Fowles & Chris Silvia
While Stata has a powerful point-and-click interface and interactive command line interface, users that rely on these exclusively are missing out on some of Stata's most powerful features. This seminar will provide novice to intermediate Stata users with insights on how to work more effectively, efficiently, and productively through an exploration of Stata's built-in programming language. Users will learn useful tips and tricks on automating repetitive tasks, accessing the macros, scalars, and matrices automatically generated by Stata for building tables of descriptive statistics and results, and how to capitalize on several popular user-written Stata add-on packages that are guaranteed to make your life as a quantitative researcher easier and more productive. We will also provide users with an overview of Stata's graphing capabilities and tools for visualizing data. A basic working knowledge of Stata is required.
Contact us (quant@ku.edu) by May 2nd if interested
|
