To determine statistical results for a literature review, you must systematically extract, evaluate, and compare key quantitative metrics like effect sizes, p-values, and confidence intervals across multiple studies.
When writing a quantitative or mixed-methods literature review, simply summarizing the authors' written conclusions is not enough. You need to critically analyze the actual data behind those claims. Here is a step-by-step guide to finding and interpreting statistical findings for your review.
1. Locate the Core Data
Instead of relying on the abstract or discussion section, head straight to the "Results" section of the paper. Look for data tables, descriptive statistics, and charts. Pay close attention to the sample size (n), mean (M), and standard deviation (SD), as these baseline figures form the foundation of the study's broader claims.
2. Identify Key Statistical Metrics
Focus on the specific numbers that measure the relationships or differences being tested. The most important indicators to record include:
- p-values: These indicate statistical significance (typically p < 0.05), telling you if the results are likely due to chance.
- Effect sizes: Metrics like Cohen’s d or Pearson’s r show the actual magnitude or strength of the relationship. Effect size is often more important than the p-value for understanding a study's real-world impact.
- Confidence intervals (CI): These provide a range of values within which the true population parameter is likely to fall, offering insight into the precision of the results.
3. Extract and Verify the Findings
Manually digging through dense statistical reporting can be exhausting, especially when dealing with complex models. If you are struggling to pinpoint the exact data points, WisPaper's Scholar QA lets you ask direct questions about a paper's methodology or results, instantly extracting the exact statistical claims and tracing them back to the specific page and paragraph. Once you locate the data, verify that the statistical tests used (such as ANOVAs, t-tests, or multiple regressions) properly align with the study's research questions.
4. Synthesize Results Across Studies
A strong literature review does not just list the statistical results of individual papers one by one. Instead, you should group the studies by their findings. Are the statistical results consistent across the literature? If five studies show a strong positive correlation but two show no significant effect, you need to highlight this discrepancy. Pointing out conflicting statistical outcomes is one of the best ways to identify valid research gaps.
5. Document Systematically
To keep track of these numbers, build a literature matrix using a spreadsheet. Create dedicated columns for sample size, statistical tests performed, effect sizes, and the primary quantitative outcomes. Logging this data as you read will make it significantly easier to spot overarching trends when it is time to write your review.
