A pair of talks on November 10 capped off the fall CARMA webcast lecture series for 2017. Dr. George Banks from the University of North Carolina-Charlotte and Dr. Todd Little from Texas Tech University both presented research during the event.
Banks opened the day with his lecture on “Questionable Research Practices.” As explained to the audience, questionable research or reporting practices (QRPs) contribute to a growing concern regarding the credibility of research in the social and natural sciences. Such practices include design, analytic or reporting practices that may introduce biased evidence, which can have harmful implications for evidence-based practice, theory development and perceptions of the rigor of science.
To assess the extent to which QRPs are actually a concern, he has conducted a systematic review to consider the evidence on QRPs. He uses a triangulation approach whereby he may be reviewing data from observations, sensitivity analyses and surveys. Through this methodology, he identified the good, the bad and the ugly. Drawing upon the findings, a series of studies have been designed and executed to evaluate the effectiveness of various types of interventions such as open data policies, results-blind reviews, study pre-registration and reviewer training. The goal of this program of research is to encourage a systematic, collegial and constructive dialogue regarding QRPs in social and natural science research.
Following Banks, Little presented his talk, “On the Merits of Parceling.” Parceling is a data pre-processing strategy by which two or more items are averaged to create a new aggregate indicator to use in both exploratory and confirmatory factor models, often referred to as latent variable modeling or structural equation modeling. First introduced by Cattell over half a century ago, the practice of parceling has been a hotly debated practice and even earned the moniker “the items versus parcels controversy.” In his lecture, he outlined the arguments both pro and con regarding the items versus parcels controversy. He concluded his talk with discussion of why the items versus parcels needn’t be one and provided compelling reasons for why parcels are highly preferred.
The entire webcast series is available online to CARMA website users. Affiliation with the University of Nebraska–Lincoln provides this access for no charge simply by signing up at https://carma.azurewebsites.net/account/register
. Members have free access to over 100 past webcasts on everything from structural equation modeling, social network analysis, item response theory and many other research methods.
CARMA webcasts resume in the spring semester on January 26 with Dr. Lisa Lambert from Georgia State University and her talk “Storytelling through Statistics,” and Dr. Daniel Beal from Virginia Tech presenting “Experience Sampling Methods.” For the entire schedule, visit the CARMA website at: http://business.unl.edu/outreach/carma/webcast-programs/