Report the normality violation directly in the methodology narrative, give the test statistic (e.g., Shapiro‑Wilk W = 0.92, p = .03), note that the assumption is breached, and then justify the switch to the Mann‑Whitney U test with a brief citation of its non‑parametric nature. Place this information in the main text, not in a footnote, because graders look for a clear chain of reasoning in the body of the chapter.
What terminology does my module guide expect when I describe a non‑parametric switch?
Most MSc psychology handbooks list “assumption testing,” “distributional checks,” and “non‑parametric alternatives” as exact headings. When you write, use those phrases verbatim; the rubric often circles “assumption of normality” and “justification for test selection.”
Start the paragraph with a sentence such as “Assumption testing revealed a violation of normality for the dependent variable.” Follow with the specific test name (“Shapiro‑Wilk test”) and the result. Then transition with “Consequently, a non‑parametric alternative—Mann‑Whitney U—was employed because it does not require normally distributed data.”
Do not replace “non‑parametric” with “distribution‑free” unless the rubric explicitly mentions the latter. The term “robust” is acceptable when describing why the Mann‑Whitney is appropriate for ordinal or skewed data, but keep the core wording aligned with the module guide.
Below is a quick reference of the exact terms that typically earn full marks:
- Assumption testing
- Normality assessment
- Shapiro‑Wilk (or Kolmogorov‑Smirnov) statistic
- Violation of normality
- Non‑parametric alternative
- Mann‑Whitney U test
- Effect size (r) for non‑parametric tests
When you match these words, the grader sees that you are speaking the same language as the assessment criteria, which reduces the chance of a “terminology” deduction.
Which technical details must I include to avoid a dead‑end on the stats side?
Beyond terminology, the statistical section needs concrete numbers and APA‑style formatting. First, report the normality test statistic with two decimal places, then the exact p‑value (e.g., p = .04). Second, give the sample size for each group (n₁ = 27, n₂ = 30) because the Mann‑Whitney calculation depends on group sizes.
Third, present the Mann‑Whitney U result, the Z‑score, and the associated p‑value. A typical APA‑style sentence looks like: “Mann‑Whitney U = 312.50, Z = ‑2.13, p = .033, r = .28.” If your software (e.g., Jamovi, JASP, or SPSS) does not output the effect size automatically, compute r = Z/√N and include it.
Finally, embed a small table that follows the rubric’s layout for statistical reporting. The table below shows the required columns and common pitfalls.
| Criterion | What the grader expects | Typical mistake |
|---|---|---|
| Assumption test | Shapiro‑Wilk W, p‑value, N | Omitting the W statistic |
| Result interpretation | Clear statement of violation (p < .05) | Only saying “data were non‑normal” |
| Test justification | Explicit link to Mann‑Whitney as non‑parametric | Using “because it seemed appropriate” |
| APA formatting | U, Z, p, r all reported with correct italics | Missing italics or rounding errors |
Notice the “Typical mistake” column; avoiding those items will keep you from losing marks on the technical front.
How can I keep the write‑up safe from similarity‑check flags?
Similarity‑detection tools compare phrasing, not ideas. Because the wording for assumption testing is standardized, you can safely reuse the exact sentences that appear in your module guide, but you must surround them with your own explanatory prose.
“The Shapiro‑Wilk test indicated a significant departure from normality (W = 0.92, p = .03), so I chose the Mann‑Whitney U test, which does not assume a normal distribution.” – a peer who passed the methodology check.
Insert a brief personal rationale after the quoted sentence, such as “Given the skewed distribution of response times, the Mann‑Whitney offered a more reliable median comparison.” This blend of required language and original commentary reduces the similarity score.
Another safeguard is to paraphrase any textbook definition you include. For example, instead of copying “The Mann‑Whitney U test compares the ranks of two independent samples,” write “The test evaluates whether the rank ordering of one group differs systematically from that of another.” The core meaning stays intact, but the phrasing becomes unique.
Finally, run your draft through a free plagiarism checker before the official proofread. If any sentence shows a similarity above 30 %, rewrite it in your own voice. This extra step prevents the editor from flagging the methodology as overly generic.
What step‑by‑step plan can I start tonight to finish the section?
Break the work into bite‑size tasks so you can make progress even with a part‑time schedule. The checklist below assumes you have already run the Shapiro‑Wilk test in your chosen software.
- Copy the exact Shapiro‑Wilk output (W, p, N) into a new Word document.
- Write a one‑sentence “Assumption testing” paragraph using the terminology list from the first section.
- Insert the Mann‑Whitney output (U, Z, p, r) directly beneath the assumption paragraph.
- Craft a justification sentence that links the violation to the choice of Mann‑Whitney.
- Format both statistical results in APA style (italics for statistical symbols, proper rounding).
- Populate the table above with your actual numbers; double‑check that each cell matches the rubric’s expectations.
- Read the paragraph aloud to ensure it flows naturally and isn’t just a string of numbers.
- Run the draft through a plagiarism checker, edit any flagged phrasing, and save the final version.
If you allocate 20 minutes to each item, the entire methodology subsection can be polished in under three hours. Start with the easiest task—copying the output—so you build momentum, then move to the more thoughtful justification.
When the proofreader returns your draft, they’ll see a clean, well‑structured paragraph that meets the rubric, includes all required technical details, uses the exact terminology, and avoids similarity‑check pitfalls. That’s the combination that keeps the methodology chapter from being sent back for revisions.