Ask ten students what a Results section should do and nine will say "show what you found." Ask their professors and you'll hear something sharper: "the Results section must present the data objectively; the Discussion may interpret it." That one discipline — presentation vs. interpretation — is the entire game, and it's why so many lab reports bleed into mush: students panic about interpreting in Results, then leave nothing original for Discussion, so both sections fail. Here's how to build a Results and Data Analysis section that separates the two jobs cleanly and makes your Discussion effortless.
First, the Division of Labor: Results vs. Discussion
- Results (and Data Analysis): what the data show — numbers, patterns, outcomes of statistical tests. Objective. No "this means…", no "surprisingly…", no "suggesting that…".
- Discussion: what the data mean — interpretation, comparison to hypothesis and literature, limitations, implications.
The rule that keeps you honest: any sentence containing "therefore," "suggests," "implies," "indicates" (about meaning), or "hypothesis" belongs in the Discussion. In Results, you are a courtroom reporter, not a prosecutor.
The Architecture: Figure First, Then Text
The strongest Results sections lead with the visual and follow with the words. Never do the reverse, and never make the reader hunt for the figure the text refers to.
The standard pattern per finding:
- One orienting sentence — what test or measurement produced this result.
- The figure or table — placed immediately after first mention.
- Two to four sentences of description — the key values, the pattern, the test statistic.
- (In data-analysis-heavy reports) the statistical result — test used, test statistic, degrees of freedom, p-value.
Tables and Figures: The 8 Rules That Separate Pros From Amateurs
- One message per table/figure. If it shows three unrelated things, split it.
- Number and title everything: "Table 1: Mean Reaction Time by Condition (n = 24)." Title above tables, below figures.
- Self-sufficiency: a reader should understand a table without reading the text — full column headers, units, sample sizes.
- Never duplicate: don't put the same data in a table and a figure. Choose the clearest format for each dataset.
- Choose deliberately: tables for exact values; figures (bar, line, scatter) for patterns and comparisons; scatterplots for relationships.
- Label axes fully with units; error bars must be defined in the caption (SD? SEM? 95% CI?).
- Figures are visual arguments — the scale shouldn't exaggerate differences; start axes appropriately but never misleadingly.
- Refer to every figure/table in the text — "As shown in Figure 2…" — and order them by appearance.
What to Write for Statistical Results
For a report with statistical analysis, report each test using the field's standard notation — no narrative fluff:
- t-test: t(22) = 3.41, p = .002
- ANOVA: F(2, 57) = 4.87, p = .011
- Correlation: r(40) = .62, p < .001
- Chi-square: χ²(1, N = 80) = 6.10, p = .014
State effect sizes when your course uses them (Cohen's d, η²) — and always note when a result was not significant (that's a finding too, not a failure): "No significant difference was found between conditions, t(22) = 0.84, p = .41."
The "Describe, Don't Interpret" Conversion Table
| ❌ Interpretation (Discussion material) | ✅ Description (Results material) |
|---|---|
| "The experimental group clearly performed better, suggesting the treatment works." | "The experimental group's mean score (M = 82.4) exceeded the control group's (M = 74.1), t(38) = 2.90, p = .006." |
| "This surprising drop indicates the stimulus lost effectiveness." | "Reaction time increased across trials from 480 ms to 612 ms." |
| "These results prove the hypothesis is correct." | "The data are consistent with the predicted direction of the effect." |
The 4-Step Writing Process for Results
- Analyze before you write: run all analyses, then decide which findings answer your research question. Not every output belongs in the report — select the load-bearing results.
- Order results by the research question, not by when you collected them. The reader should see your question's answer assembled in a logical sequence.
- Draft text for one finding at a time: orient → figure → describe → statistic. Repeat.
- Do the "therefore audit": scan your draft and delete or relocate every interpretive word into the Discussion file. If your Results has no Discussion-ready sentences left, you've nailed the discipline.
Common Mistakes That Cost Points
- Number-dumping: paragraphs of "the mean was 5.2, then 6.1, then…" with no figure and no orienting sentence
- The "accidental Discussion": letting adjectives sneak in ("dramatically," "clearly," "surprisingly")
- Reporting outliers without explanation (note them and how you handled them)
- Repeating the methods: Results describes findings, not procedures — one sentence of orientation max
- Raw data in the body: raw logs live in appendices; the Results section shows processed, summarized findings
Conclusion
The Results and Data Analysis section wins on discipline: present what the data show, keep every judgment for the Discussion, and let figures and tables carry the heavy lifting. Commit the conversion table to memory, run the "therefore audit" before submission, and your lab reports will finally have the crisp objectivity graders are hunting for — and your Discussion section will write itself from the interpretations you wisely saved. Next lab: let the data talk first, then you'll have something worth saying.
Put it into practice
Choose one idea from this guide and apply it in your next study block. Track what changes, then refine your process based on the evidence—not the feeling of familiarity.