The second Consortium for the Advancement of Research Methods and Analysis (CARMA) webcast lectures took place on October 20 with a pair of talks. Dr. Larry Williams, director of CARMA and Donald and Shirley Clifton Chair of Survey Science and Professor of Management, presented “Condition 9 and 10 Tests of Model Confirmation with SEM Techniques,” and Dr. Richard Landers, associate professor of psychology at Old Dominion University, spoke on “Creating Datasets with Social Media.” 
 
The lecture by Williams presented the unique contribution of the “conditions” required for appropriate confirmatory inference with the path and latent variable models. He discussed the importance of James et al.’s Condition 9 and 10 tests, systematically reviewed 14 years of studies using SEM in leading management journals and reanalyzed results based on new techniques that extend James et al. (1982). He concluded with suggestions for improved Condition 9 and 10 assessments.
 
Next, Landers gave his talk, “Creating Datasets with Social Media.” He discussed the breadth of publicly available behavioral data from social media, citing two major challenges to collecting such data and analyzing it meaningfully. He noted that concerns include external validity issues and addressing the technical skills needed to actually collect such data. To address both of these concerns, he first described data source theories, a formal theoretical approach to understanding generalizability from big data-based samples. Then, Landers demonstrated how to create a simple dataset within this approach from Facebook with very little programming required – less than five lines of easy-to-read code in R.
 
The entire webcast series is available online and you are encouraged to become a CARMA website user. Affiliation with the University of Nebraska-Lincoln provides this access for no charge simply by signing up: https://carma.azurewebsites.net/account/register. Once a member, you will have free access to more than 100 past webcasts on everything from structural equation modeling, social network analysis, item response theory and many other research methods.  Published: November 15, 2017