Quantitative Analysis of Pre-Existing Data

In short:  Pre-existing data can be found in publicly-available datasets or from labs where you’re serving as an RA. The challenge comes in crafting new research questions using variables and a sample you didn’t get to choose. If your data comes from a lab, the question of authorship can be a tricky one. Just make sure all parts of your IP submission were written by you. No coauthors are permitted on your IP, even if results from your project will later be published with a team. Similarly, already-published articles may not be submitted for your IP.  TC’s Institutional Review Board (IRB) may need to be consulted if identifiable human subjects data is used, so be sure to speak to your lab leader. 

Key Components

• Literature review
• Research questions
• Valid instruments & measures
• IRB approval
• Proper sample size
• Statistical tests 
• Results & Discussion

The whole story:  A quantitative analysis of pre-existing data involves processing and statistically analyzing previously collected numerical data (e.g., surveys, experiments) to investigate a clearly defined and falsifiable research question or hypothesis. Your research questions should go beyond the original purpose of the previously-collected dataset. You should conduct a novel analysis that produces additional insights not investigated in the original study. 

The process begins with an appropriately comprehensive review of existing empirical literature on the chosen topic, which should conclude with a review of remaining gaps in the literature and/or future directions still needed. This initial literature review should be written in a manner that serves to justify the value of the current study, because it might be a valuable addition to the existing knowledge base or could inform future clinical interventions.

The student must next pose a specific, testable set of research questions or hypotheses that are answerable using the data in the pre-existing dataset. Again, these research questions and hypotheses should go beyond the original purpose of the previously collected dataset. A high-quality quantitative research question investigates whether a meaningful relationship exists between a narrow set of clearly defined variables that can be measured using standardized methods such as Likert-type questionnaires, behavioral observations, or experimental tasks. 

To this end, the student should next identify the research instruments (e.g., scales, tests, experiments, etc.) in their pre-existing dataset that can be applied to answer their questions or hypotheses. These instruments should have adequate reliability and validity. 

The student must obtain approval of an “exempt” IRB protocol for a secondary analysis of existing data. This IRB protocol should describe their study and describe their approach to securely handling data in the secondary dataset. Students should design their secondary study in the lowest risk manner possible to answer their research questions. 

The chosen dataset should have a sample size large and robust enough to support meaningful statistical analysis, but this can include smaller sizes that would qualify as a “pilot study.” Power analysis should be used to aid in identifying a minimally-viable sample size (e.g., using G*Power or a similar software tool). 

Statistical techniques used must suit the data and question, and quantitative statistical software, such as SPSS or R software, should be used to conduct inferential statistical analyses in a rigorous manner. The student must check assumptions (e.g., normally distributed or skewed data) and explain all statistical choices made and steps taken. Results must be clearly explained and linked back to the research question.

The Discussion should provide a nuanced account of how results add to the existing empirical knowledge base, inform future targets for clinical intervention, or otherwise add value to the field of clinical psychology. Limitations should be acknowledged and possible future directions for follow-up research should be delineated. The final integrative project should follow a clear structure (e.g., abstract, introduction, methods, results, discussion, references) and provide well-organized tables or figures as applicable. All references should be properly listed and all writing should be in APA style.

To learn more:  APA’s Journal Article Standards for Quantitative Research (https://apastyle.apa.org/jars/quantitative), including a detailed and useful outline (https://apastyle.apa.org/jars/quant-table-1.pdf).

Relevant Courses at TC:   

• Research Methods in Clinical Psychology (CCPX)
• Advanced Research Methods in Clinical Psychology (CCPX)
• Programing for Psych Research (CCPX)
• Any of the many Statistics courses offered by HUDK 

Please see our Research Methods Concentration for a full list of courses