What Quantitative Findings Look Like
Quantitative findings deal with numbers — means, percentages, correlations, regression coefficients, p-values, and effect sizes. The goal is to be precise, objective, and replicable. You report what the numbers say; you save the why for Chapter 5.
Basic Structure for Quantitative Findings
- Descriptive statistics — sample demographics, means and SDs of key variables.
- Preliminary analysis — data cleaning, assumption testing (normality, outliers, multicollinearity).
- Results by research question — present each research question in order.
- Statistical test results — tables with coefficients, significance levels, and effect sizes.
- Unexpected findings — surprising patterns or null results.
- Summary of findings — brief restatement of key numbers.
Descriptive Statistics: Sample Characteristics
Always start by describing who or what you studied. Use a table for demographics.
Table 4.1 — Demographic Characteristics of Survey Respondents (N = 342)
| Characteristic | n | % |
|---|---|---|
| Gender | ||
| Female | 198 | 57.9 |
| Male | 136 | 39.8 |
| Non-binary / Prefer not to say | 8 | 2.3 |
| Age (years) | ||
| 18–24 | 45 | 13.2 |
| 25–34 | 187 | 54.7 |
| 35–44 | 72 | 21.1 |
| 45+ | 38 | 11.1 |
| First-generation college student | ||
| Yes | 156 | 45.6 |
| No | 186 | 54.4 |
How to write about it: "A total of 342 graduate students completed the survey. The sample was predominantly female (57.9%) and aged 25–34 (54.7%). Nearly half (45.6%) identified as first-generation college students."
Descriptive Statistics: Key Variables
Report means and standard deviations for each main variable.
Table 4.2 — Descriptive Statistics for Continuous Study Variables (N = 342)
| Variable | M | SD | Possible Range | Observed Range |
|---|---|---|---|---|
| Academic self-efficacy (1–5) | 3.67 | 0.82 | 1 – 5 | 1.2 – 5.0 |
| Social integration (1–5) | 3.42 | 0.91 | 1 – 5 | 1.0 – 5.0 |
| Persistence (0/1) | 0.74 | 0.44 | 0 – 1 | 0 – 1 |
Preliminary Analysis (Assumption Testing)
Before running inferential tests, check that your data meet the assumptions of the chosen statistical test. Report:
- Missing data handling — how many cases were excluded and why.
- Outliers — how identified (e.g., z-scores > |3.29|) and whether retained or removed.
- Normality — skewness and kurtosis values (acceptable ranges: skew < |2|, kurtosis < |7|).
- Linearity (for regression) — scatterplot inspection.
- Homoscedasticity — residual plot.
- Multicollinearity — VIF and tolerance values.
- Independence of errors — Durbin-Watson statistic.
Example paragraph: "Of the 360 returned surveys, 18 were excluded due to missing data on the dependent variable, leaving N = 342 for analysis. Skewness (–0.42 to 0.81) and kurtosis (–0.61 to 1.32) values were within acceptable ranges. Variance inflation factors ranged from 1.08 to 1.42, indicating no multicollinearity concerns."
Statistical Test Results: Tables and Figures
Always include the test name, degrees of freedom, test statistic, p-value, and effect size where possible.
Table 4.3 — Multiple Regression Predicting Persistence (N = 342)
| Predictor | B | SE(B) | β | t | p |
|---|---|---|---|---|---|
| (Intercept) | 0.24 | 0.08 | — | 3.00 | .003 |
| Academic self-efficacy | 0.31 | 0.06 | .34 | 5.17 | <.001 |
| Social integration | 0.18 | 0.05 | .22 | 3.60 | <.001 |
| First-gen status | –0.09 | 0.04 | –.11 | –2.25 | .025 |
Note. R² = .28, F(3, 338) = 43.81, p < .001.
Reporting Unexpected Quantitative Findings
If a hypothesized relationship was not significant, say so plainly. If a covariate behaved oddly, note it. Honest reporting strengthens credibility — and gives you material for your discussion.
Common Mistakes
- Interpreting results inside Chapter 4 ("This shows that…").
- Reporting only p-values without effect sizes.
- Forgetting to align results with the order of your research questions.
- Dropping raw SPSS / R outputs into the chapter without narrative.
Frequently Asked Questions
Do I need to report every test I ran?
Report only tests that address your research questions. Exploratory or post-hoc analyses can go in an appendix or be mentioned briefly as unexpected findings.
What about effect sizes?
Always include them. Cohen's d, η², R², and odds ratios convey practical significance that p-values alone cannot.
Continue to: Part 2: Qualitative Findings →