1. SECONDARY DATA ANALYSIS

2. Overview  Background  Components  Examples  Challenges  Opportunities

3. What is it? (Hint: You’re already doing it)  Secondary data analysis  Analyzes data from an existing data set Data you did not collect yourself (colleague, collaborator, professor) Data from a study focused on something else (primary analyses/papers already complete) Data collected for some other purpose (medical records, transcripts, etc.)  Also called “primary analysis of existing data” or “archival data analysis” Distinguishes from merely re-testing the same hypotheses

4. Data Sets: Examples  Inter-university Consortium for Political and Social Research (ICPSR)  http://www.icpsr.umich.edu/icpsrweb/ICPSR/access/su bject.jsp http://www.icpsr.umich.edu/icpsrweb/ICPSR/access/su bject.jsp  Catalogue of many public data sources  Midlife in the United States (MIDUS)  http://www.midus.wisc.edu/ http://www.midus.wisc.edu/  National longitudinal study of health and well-being

5. Components 1. Permission  Often requires a “data usage agreement” Rules about which analyses will be conducted, authorship, timeliness  May require IRB approval 2. Data dictionary  Protocol, measures, descriptive statistics  Quality varies: Non-existent, paper files, Excel, PDF 3. Data file  SPSS, Excel, SAS

8. Challenges  Upfront investment: Varying quality of data dictionaries and collaborators  Imperfect measures  You might want a quality measure of extraversion, and all they have is a single item on optimism  Politics: Some psychology departments quantify science in terms of studies-per-paper rather than papers-per-study  may be viewed as lesser science, or may require theses/dissertations to use “original” data collection

9. Opportunities  Upfront investment = delayed gratification  Efficient on many levels  Implementation, publishing, funding  Significance  High-quality studies  May begin by asking the PI, “What is the most important finding in the data set you’ve never had time to publish?”  Can be symbiotic (mutually beneficial)  Own posters/presentations/publications, PI is happy for added productivity (publications are good, helps with funders)

10. Salzman study as of 9/24/15 Data Set PI MenteesExternal Collaborators Failed Aim 1 Article Manuscript Submitted Manuscript Manuscript Submitted Manuscript Funded Grant Manuscript Submitted Masters Thesis Manuscript Submitted Manuscript Funded Grant Also: 16 conference posters/presentations, 3 promotions, 3 students advancing on to higher degree programs, 4 students who were able to attend a national conference