Convergence: Connecting Levels of Analysis in Psychological Science
 In the past, our field harbored distinct, and often competing, schools of thought that tackled different problems and produced findings that often appeared to diverge. Today, investigators attack shared problems at complementary levels of analysis and produce results that converge. Studies of people in a social world; mental systems of cognition and emotion; and biological mechanisms of the genome and the nervous system interconnect and yield an integrated psychological science. The APS 23rd Annual Convention displays, and celebrates, these advances in our field.


Integrative Data Analysis: Applications Across Different Data Types

Thursday, May 24, 2012, 1:30 PM - 5:20 PM
Chicago Ballroom IX

Richard P. Moser Chair: Richard P. Moser
National Cancer Institute

Integrative data analysis (IDA) is a general term for a set of analytic techniques derived from combining or linking independent data sets together and analyzing them as a complete set. This is different from meta-analysis in the sense that one analyzes the actual data in IDA, not the statistical summaries of those data. IDA is a cost-effective way to do science and has the potential to move areas of science forward rapidly by building a cumulative knowledge base. It is an extremely topical issue given the unprecedented access to data that is now afforded to all researchers through cyberinfrastructure (i.e., Internet-based research environments) and a push from the Federal government to make data more accessible.

This 4-hour workshop will provide a general overview of the pertinent issues involved with IDA, demonstrate three applied guided examples utilizing different types of data, and discuss federal funding opportunities to support IDA methodology. Statistical code and related output will be provided to workshop participants so that they can follow along with each example.

Workshop Objectives:

1. Learn about the conceptual and analytic issues involved with integrative data analysis

2. Observe applied guided examples of the types of integrative data analyses that can be done

3. Apply techniques learned to a prescribed dataset during a workshop

Download the List of References

Patrick J. Curran

Longitudinal Measurement Modeling in Integrative Data Analysis
Patrick J. Curran
University of North Carolina at Chapel Hill
One of the most significant challenges in longitudinal integrative data analysis (IDA) is to establish valid and reliable measurement that is sensitive to potential structural differences across group (e.g., gender, treatment) and developmental period (e.g., childhood to adolescence to adulthood). This presentation will highlight these issues by demonstrating the fitting of measurement models to 18 binary items assessing internalizing symptomatology drawn from three separate studies covering a developmental period of 11-34 years of age. Recommendations for the use of these techniques in other settings will be provided.

Michael Larsen

Record Data Linkage
Michael Larsen
The George Washington University
Record linkage, or exact matching, involves finding the individuals that have records in two or more databases and merging information to create richer, micro data records for these individuals. The linked data can enable analysis that would not be possible using the databases separately. Strict exact matching on several variables ensures that few false matches (false positives) are included on the merged data base. Some matches might be missed, however, due to missing data, typographical errors, and alternative versions of information. Lose matching criteria generally avoid missing as many matches, but create false matches, which can distort analyses. This talk will illustrate steps in one or more record linkage applications and discuss common concerns of researchers interested in record linkage and the analysis of linked files.

Daniel J. Bauer

Harmonizing Classification Variables in Integrative Data Analysis
Daniel J. Bauer
University of North Carolina at Chapel Hill
This talk will discuss strategies for harmonizing classification variables when classification criteria differ between studies or evolve over time. For instance, DSM criteria for alcohol and substance use disorders have undergone revision, creating the potential for measurement differences in diagnostic classifications across studies conducted at different points in time or within a given longitudinal study over time. The task is then to harmonize the classifications to a common criteria set or target definition. This talk presents and contrasts two approaches for scoring classification variables to a target definition, latent class/profile analysis, and multiple imputation.

Sierra P. Moser

Co-Author: Sierra Bainter, University of North Carolina at Chapel Hill

Subject Area: Methodology

Email Bookmark and Share