Open StatClinic →
⚠ Research Methodology

7 Common Statistical Mistakes
That Ruin Medical Research Papers

🕑 26 min read 📅 July 2026 ✅ Peer-reviewed content 📚 3200+ words
S
StatClinic Editorial Team Statistical content for medical researchers and clinicians
Statistical errors are more common in published medical research than most clinicians realise. A 2022 methodological audit of 250 original research papers across five high-impact clinical journals found that 64% contained at least one identifiable statistical flaw — ranging from minor reporting omissions to fundamental errors that invalidated the primary conclusion. These mistakes are not confined to student theses or low-tier publications. They appear in RCTs, cohort studies, and systematic reviews in journals with impact factors above 10. Understanding the seven most dangerous statistical mistakes — and learning to prevent them before data collection begins — is one of the highest-return investments any medical researcher can make.

The Seven Mistakes at a Glance

Each mistake described in this guide represents a category of error identified repeatedly in published methodological audits and statistical peer reviews. They are ordered roughly from most to least frequently encountered in clinical research submissions.

QUICK REFERENCE — 7 STATISTICAL MISTAKES IN MEDICAL RESEARCH
01
Selecting the wrong statistical test for the data type
02
Skipping the normality check before parametric tests
03
Starting a study without a proper sample size calculation
04
Misusing and misinterpreting the p-value
05
Overfitting regression models to small datasets
06
Treating questionnaire and Likert data incorrectly
07
Misinterpreting the meaning of results
+
Prevention checklist and FAQ follow each section

Mistake 1: Selecting the Wrong Statistical Test

01
Mistake #1
Wrong Statistical Test Selection
Critical Impact Very Frequent

Applying a statistical test that does not match the data type, the number of groups, or the distributional assumptions of the outcome variable. The most common versions: using a t-test on ordinal or non-normal data, applying chi-square when expected cell frequencies are below 5, using Pearson correlation on a non-linear relationship, or running one-way ANOVA without checking the sphericity assumption for repeated measures.

A rheumatology study compares post-operative pain intensity — measured on a 0–10 visual analogue scale (VAS) — between two NSAID regimens using an independent samples t-test. The VAS is ordinal: the difference between a score of 3 and 4 is not necessarily the same as between 7 and 8, and in an acute pain population, scores cluster at the low end (floor effect). The correct test is the Mann-Whitney U test. Running a t-test on this data inflates the Type I error rate and produces an invalid p-value. The published conclusion — "Regimen A provides significantly greater analgesia" — is not supported by the appropriate analysis.

A wrong test produces a wrong p-value. The conclusion — significant or non-significant — may be incorrect in either direction. Statistical reviewers at major journals now use systematic checklists, and test-selection errors are a documented leading cause of peer-review rejection and post-publication corrections. A 2020 audit of 160 surgical RCTs found that 22% had used a parametric test on clearly non-normal outcomes without justification.

  1. Before choosing any test, classify your outcome variable explicitly: continuous (interval/ratio), ordinal, or categorical (nominal).
  2. Ask: Are groups independent or paired/matched? How many groups? Are there confounders requiring adjustment?
  3. Run a normality check (Shapiro-Wilk) before deciding between parametric and non-parametric.
  4. Use a validated statistical decision tree — the StatClinic test selection guide covers every branch of this decision.
  5. State your test selection rationale explicitly in the Methods section, including any assumptions checked.
✗ Wrong Independent t-test on VAS pain scores
(ordinal, non-normal, floor-effected)
Assumes normality and interval-level spacing — both violated here
✓ Correct Mann-Whitney U test
(non-parametric, 2 independent groups)
No normality assumption; appropriate for ordinal and skewed continuous data

Mistake 2: Ignoring Normality Assumptions

02
Mistake #2
Skipping the Normality Check
Critical Impact Very Frequent

Applying the t-test, ANOVA, Pearson correlation, or linear regression without ever checking whether the outcome variable (or regression residuals) is approximately normally distributed. Many researchers assume that clinical data is normal by default, or perform the analysis first and add a normality check only if a reviewer requests it. This is methodologically backwards: the normality check must precede test selection, not follow it.

A hepatology team compares serum alanine aminotransferase (ALT) between alcoholic hepatitis patients and healthy controls using an independent t-test. ALT is structurally right-skewed — the majority of values cluster near the normal range (20–50 U/L) while a minority spike into the thousands during acute hepatic injury. No normality assessment is reported in the Methods. Shapiro-Wilk on the hepatitis group returns W=0.68, p<0.001. The t-test is invalid. The correct approach is Mann-Whitney U, with results reported as median (IQR) rather than mean ± SD.

Parametric tests applied to substantially skewed data produce incorrect p-values — typically underestimating variance in the tails, generating spuriously narrow confidence intervals and falsely low p-values. A conclusion of statistical significance may be an artefact of violating the normality assumption rather than a genuine treatment effect. A 2019 systematic review of 112 cardiology RCTs found that 48% applied parametric tests without reporting normality assessment.

  1. Run the Shapiro-Wilk test (n < 50) or Kolmogorov-Smirnov with Lilliefors correction (n > 50) on every outcome variable before selecting a test.
  2. Inspect a histogram and a Q-Q plot in SPSS — visual assessment is essential, especially in large samples where Shapiro-Wilk is oversensitive.
  3. Check the skewness statistic: values >1 or <−1 indicate meaningful skew warranting non-parametric tests.
  4. For variables known to be structurally non-normal (ALT, CRP, troponin, creatinine, LOS, costs), use non-parametric tests by default.
  5. Document in your Methods: "Normality was assessed using the Shapiro-Wilk test. Non-normally distributed variables were analysed using non-parametric tests and reported as median (IQR)."
✗ Wrong t(48) = 2.31, p = 0.025
Mean ALT: 142 vs 28 U/L
Normality not checked; t-test applied to severely right-skewed biomarker
✓ Correct U = 418, z = 3.84, p < 0.001
Median ALT: 124 (IQR 67–289) vs 26 (IQR 19–34) U/L
Non-parametric test; correct central tendency and spread for skewed data

Mistake 3: Inadequate Sample Size

03
Mistake #3
Starting a Study Without Sample Size Calculation
Critical Impact High Frequency

Beginning data collection without a prospective sample size calculation — or performing the calculation after data collection to justify the numbers available. Underpowered studies cannot reliably detect clinically important differences even when they exist (Type II error). The resulting false-negative conclusion can waste years of research, mislead systematic reviews, and delay effective treatments from reaching patients. The complementary error — vastly over-recruiting — wastes resources and unnecessarily exposes additional patients to study procedures.

A surgical team designs an RCT comparing surgical site infection (SSI) rates between standard wound care (expected 30%) and a new antimicrobial dressing (target reduction to 15%). They enrol n=20 per group. A properly powered study (two-sided chi-square, α=0.05, power=0.80) requires n=82 per group for this effect size. After analysis, the study finds SSI rates of 28% vs 19% (p=0.07). The team concludes: "No statistically significant benefit was demonstrated." This conclusion is unwarranted. The study was fatally underpowered — it had less than 30% power to detect a 15-percentage-point difference at this sample size. The dressing may genuinely reduce infections, but the study was too small to know.

Underpowered studies generate unduly wide confidence intervals that straddle the null, leading to false acceptance of no effect. These null results accumulate in systematic reviews, providing false scientific weight to ineffectiveness. A 2021 analysis of 119 orthopaedic RCTs found that 27% were underpowered to detect their primary outcome at 80% power. Post-hoc power analysis — calculating power after the fact to explain a non-significant result — is methodologically meaningless and is rejected by all major statistical bodies.

  1. Calculate sample size before data collection begins — always, without exception.
  2. Specify the expected effect size (from pilot data or published literature), α (typically 0.05), desired power (typically 0.80 or 0.90), and the appropriate test.
  3. Add a 10–20% inflation factor for expected dropout, non-compliance, and missing data.
  4. Report the complete formula and all inputs in the Methods section: "Sample size was calculated using [formula/tool], assuming [effect size], α=0.05, power=0.80, with 15% attrition adjustment, giving n=X per group."
  5. If a study is already complete and underpowered, report the 95% CI, acknowledge the limitation explicitly, and do not accept the null hypothesis based on p > 0.05 alone.

Mistake 4: Misusing the P-value

04
Mistake #4
P-value Misuse and Misinterpretation
Critical Impact Extremely Frequent

1. Treating p > 0.05 as "no effect." Absence of statistical significance is not evidence of absence of an effect. A non-significant result in a small sample may simply reflect insufficient power. 2. Treating p < 0.05 as clinical proof. Statistical significance and clinical importance are entirely different concepts. A large RCT can produce p < 0.001 for a difference that is too small to matter to any patient. 3. P-hacking: running multiple subgroup analyses, outcomes, or models and reporting selectively only those achieving p < 0.05, without correcting for multiple testing.

A cardiologist compares LDL reduction between atorvastatin 20 mg and 40 mg in n=30 per group. Mean difference: 8.2 mg/dL (95% CI: −0.4 to 16.8 mg/dL; p=0.062). The paper concludes: "There was no significant difference in LDL reduction between the two doses." The CI shows that the true effect could plausibly be as large as 16.8 mg/dL — a clinically important reduction. Declaring no effect based on p=0.062 alone is unjustified and misleading.

A researcher tests a drug's effect across 8 pre-specified subgroups (age >60, male sex, hypertension, diabetes, BMI >30, smoking, CKD, prior MI) without correction for multiple comparisons. One subgroup (males >60 with diabetes) achieves p=0.031. The discussion highlights this subgroup as a "clinically important finding." With 8 simultaneous tests at α=0.05, the probability of at least one false positive by chance alone is 1 − (0.95)⁴ ≈ 34%. Without Bonferroni correction (p < 0.006 per test), this result is unreliable.

P-value misuse is the statistical engine of the reproducibility crisis. Ioannidis's landmark 2005 analysis argued that the majority of published medical research findings may be false, with selective reporting and p-value misuse among the primary drivers. Journals including JAMA, The BMJ, and the New England Journal of Medicine have since updated their statistical reporting requirements to mandate effect sizes, confidence intervals, and pre-registration. Papers without these elements face increasingly difficult peer review.

  1. Always report the effect size alongside the p-value: Cohen's d, OR, RR, η², Cramér's V, or Pearson r as appropriate.
  2. Report 95% confidence intervals for every key estimate — the CI conveys what the p-value cannot: the direction and plausible magnitude of the effect.
  3. Pre-register your primary outcome and analysis plan in a trial registry (ClinicalTrials.gov, ISRCTN) or on OSF before data collection.
  4. Apply Bonferroni correction (p < α/k) or Benjamini-Hochberg FDR correction when conducting multiple simultaneous comparisons.
  5. Never write "no difference was found" when p > 0.05 in an underpowered study. Write instead: "The study was not sufficiently powered to detect a clinically important difference; no firm conclusion regarding equivalence can be drawn."

Mistake 5: Overfitting Regression Models

05
Mistake #5
Overfitting Regression Models to Small Datasets
High Impact Frequent in Clinical Prediction

Including too many predictor variables for the sample size available in a multivariable regression model. An overfitted model learns the random noise in your specific dataset rather than the true underlying signal — producing inflated apparent performance metrics (R², AUC, sensitivity) that collapse dramatically when the model is tested in a new patient population. Overfitting is also caused by data-driven stepwise variable selection: iteratively adding or removing variables based on p-values, which capitalises on chance associations in the dataset.

A critical care team develops a logistic regression model to predict in-hospital mortality in sepsis patients. They have n=80 patients with 22 deaths (the outcome events). They include 18 predictors: age, sex, BMI, diabetes, hypertension, CKD, lactate, creatinine, bilirubin, temperature, heart rate, MAP, GCS, SOFA score, antibiotic delay, fluid volume, vasopressor use, and infection source. The model achieves AUC=0.91 in the training set. Two independent validation attempts achieve AUC=0.61 and 0.58 — barely better than random. The Events Per Variable (EPV) ratio was 22 ÷ 18 = 1.2 — catastrophically below the recommended minimum of 10. The model overfit massively to noise.

An overfitted clinical prediction model, when deployed in real-world practice, will perform unpredictably across different patient populations and clinical settings. Decisions about ICU escalation, early discharge, or treatment initiation based on such a model can directly harm patients. Overfitted models published without external validation have contributed to flawed clinical decision-support tools that failed at the bedside. Peer reviewers increasingly demand bootstrap validation or prospective external validation before publication.

  1. Apply the EPV rule: Events ÷ Number of Predictors ≥ 10 (conservative: aim for ≥ 20).
  2. Pre-specify your predictor variables based on biological plausibility and prior literature — never use stepwise, backward, or forward selection driven by p-values.
  3. Perform internal validation: bootstrap resampling (500+ iterations) or k-fold cross-validation to obtain corrected performance estimates.
  4. Perform external validation in an independent dataset before publishing a prediction model.
  5. Report calibration (Hosmer-Lemeshow test or calibration plot) alongside discrimination (AUC/C-statistic). A high AUC with poor calibration indicates an overfitted model.
✗ Wrong 18 predictors, n=80, 22 events
EPV = 1.2 → AUC = 0.91 (training)
Severe overfitting; AUC collapses to 0.58–0.61 in validation
✓ Correct 2 predictors (SOFA + lactate)
EPV = 11 → Bootstrap-corrected AUC = 0.78
Parsimonious model; performance estimate adjusted for optimism

Mistake 6: Incorrect Questionnaire and Likert Scale Analysis

06
Mistake #6
Treating Ordinal Questionnaire Data as Continuous
High Impact Very Frequent in Thesis Research

1. No reliability check: Applying a questionnaire without reporting Cronbach's alpha. 2. Parametric tests on individual Likert items: Treating a 5-point Likert item as if it were measured on a continuous interval scale and applying a t-test. 3. Summing unreliable items: Computing a composite score from items that do not measure the same construct (low internal consistency). 4. Skipping factorial validity: Assuming a multi-item scale measures a single construct without checking dimensionality via exploratory or confirmatory factor analysis.

A physiotherapy study evaluates patient satisfaction using a 5-item questionnaire (each item: 1=Strongly Disagree to 5=Strongly Agree). The researcher sums the five items into a composite score (range 5–25) and runs an independent t-test to compare two treatment regimens. No Cronbach's alpha is reported. SPSS analysis reveals α=0.43 — well below the accepted minimum of 0.70. The five items are measuring different and inconsistent constructs. Summing them is therefore statistically invalid. The t-test comparison of the composite is meaningless — the composite is noise. The paper was rejected during peer review on this specific methodological ground.

An unreliable questionnaire contains excessive random measurement error. When t-tests or ANOVA are applied to such data, they may return statistically significant results driven by noise rather than genuine group differences — or conversely, fail to detect genuine differences because the outcome measure is too imprecise. Research conclusions built on unvalidated or misanalysed PROMs cannot be trusted or replicated. Journal editors in nursing, rehabilitation, psychiatry, and primary care have significantly tightened requirements for PROM reporting following systematic identification of this problem.

  1. Use an existing validated questionnaire (SF-36, EQ-5D, HADS, PHQ-9, ODI, AUDIT) where possible — validity and reliability are pre-established.
  2. If using a new questionnaire, report Cronbach's α ≥ 0.70 as a minimum threshold for acceptable internal consistency; include this in the Methods section.
  3. For individual Likert items, use non-parametric tests: Mann-Whitney U (2 groups), Kruskal-Wallis (3+ groups), Spearman correlation.
  4. For validated composite scores that are approximately normally distributed, parametric tests may be defensible — check and report the distribution.
  5. Report dimensionality: cite the original validation study or run exploratory factor analysis if using the scale in a new population.
✗ Wrong Composite score (5 items summed)
t(58)=2.14, p=0.036
No Cronbach's α reported
α=0.43 makes composite invalid; t-test result is meaningless
✓ Correct PHQ-9 composite (α=0.86)
Approx. normal; t(58)=2.14, p=0.036
or Mann-Whitney U for single items
Validated tool with confirmed reliability; appropriate test for data type

Mistake 7: Incorrect Interpretation of Results

07
Mistake #7
Misinterpreting What the Results Actually Mean
High Impact Common in Discussion Sections

Even when analysis is technically correct, the interpretation in the Discussion section can undermine the entire paper. The four most dangerous interpretation errors: (1) inferring causation from association in observational data; (2) treating statistical significance as clinical significance; (3) reporting an odds ratio as if it were a relative risk when the outcome is common; (4) interpreting a weak correlation coefficient (e.g., r=0.28) as evidence of a "strong" or "meaningful" association.

A cross-sectional ICU study finds that serum albumin correlates with 28-day mortality: r=−0.39, p=0.001. The Discussion concludes: "Low albumin causes increased ICU mortality. Albumin supplementation should therefore be considered as a therapeutic strategy." Serum albumin is a marker of critical illness severity, not a modifiable cause of death in most patients. Confounding by illness severity is entirely ignored. Dozens of RCTs of albumin infusion based on exactly this logic have found no mortality benefit or harm in certain groups.

A case-control study finds OR=2.8 for a dietary risk factor. Outcome prevalence in controls is 38%. The paper states: "Exposure to this dietary factor more than doubled disease risk." With high baseline prevalence (38%), the OR substantially overestimates the relative risk. Using the Zhang & Yu conversion formula: RR ≈ OR ÷ [(1−P₀) + (P₀ × OR)] = 2.8 ÷ [(0.62) + (0.38 × 2.8)] ≈ 1.68. The risk increased by 68%, not 180%. This distinction has direct implications for policy, NNT calculations, and clinical counselling.

Interpretation errors in high-profile papers directly shape clinical practice. Physicians reading a well-designed study that incorrectly concludes causation may implement interventions that are ineffective or harmful. Misreported ORs in case-control studies of common diseases have been cited in guidelines and systematic reviews, inflating perceived risk magnitudes. A 2023 umbrella review of cardiovascular risk factor studies found that >30% of case-control papers reporting ORs for common outcomes (prevalence >15%) did not acknowledge the OR-RR discrepancy in their interpretation.

  1. Never use the word "causes" from observational data unless you have performed a formal causal analysis (propensity score matching, instrumental variable, Mendelian randomisation, or difference-in-differences).
  2. In case-control studies where outcome prevalence exceeds 10%, convert OR to RR using the Zhang & Yu formula and report both clearly.
  3. Compare your effect size to the published Minimum Clinically Important Difference (MCID) for your outcome. Acknowledge explicitly if the statistically significant effect does not reach the MCID.
  4. Apply the standard correlation benchmarks: r=0.1–0.3 = small, r=0.3–0.5 = moderate, r > 0.5 = large (Cohen, 1988). Do not describe a weak correlation as "strong" or "clinically meaningful" without empirical justification.
  5. State design limitations clearly in the Discussion: "As a cross-sectional study, this design cannot establish temporality or causation. Residual confounding by unmeasured variables cannot be excluded."

Journal Publication Consequences

Each of these mistakes has a documented track record of consequences at the peer review and post-publication stages. Understanding what reviewers look for helps you pre-empt rejection.

# Mistake Reviewer Response Severity
01 Wrong test selection Major revision: reanalysis with correct test required; results may change significantly Critical
02 Normality not checked Major revision: normality assessment and potentially reanalysis required Critical
03 Underpowered study Rejection or major revision; conclusion of "no effect" challenged; more recruitment required Critical
04 P-value misuse Major revision: effect sizes + CIs mandated; selective reporting may trigger editorial investigation Critical
05 Overfitting regression Rejection: external validation or cross-validation required before prediction model is publishable High
06 Questionnaire errors Major revision: reliability stats required; potential reanalysis with non-parametric tests High
07 Misinterpretation Major revision: causal language must be removed; OR/RR distinction must be addressed High
Post-Publication Statistical errors are among the leading documented causes of post-publication corrections and retractions tracked by Retraction Watch. Tools like GRIM, SPRITE, and StatCheck are now used by journals and readers to forensically detect impossible means, mismatched SDs, and recalculated p-values that do not match reported test statistics. Prevention before submission is incomparably less costly than a correction or retraction.

Practical Prevention Checklist

Use this checklist as a pre-submission quality control gate. Every item below corresponds to one of the seven mistake categories. Clear all items before submitting to any journal.

PRE-SUBMISSION STATISTICAL QUALITY CHECKLIST
📋 Before Data Collection
Sample size calculated a priori with specified effect size, α, power, and attrition adjustment
Primary outcome pre-specified and analysis plan registered or documented before data collection begins
Statistical test pre-selected based on the data type and number of groups — confirmed appropriate before collecting data
Questionnaire reliability confirmed: using a validated tool, or planning to report Cronbach's α for a new instrument
📊 Before Running Statistical Tests
Normality assessed for every continuous outcome variable using Shapiro-Wilk (n < 50) or K-S with Lilliefors (n > 50), plus histogram and Q-Q plot
Parametric vs non-parametric confirmed for each analysis based on normality results
EPV checked for any regression model: outcome events ÷ predictors ≥ 10
Cronbach's alpha calculated for any composite questionnaire score (α ≥ 0.70 required)
📝 During Analysis and Reporting
Effect size reported alongside every p-value: Cohen's d, OR, RR, η², or Pearson r as appropriate to the test
95% confidence intervals reported for all primary and secondary outcome estimates
Multiple comparisons corrected: Bonferroni or Benjamini-Hochberg applied when performing > 1 simultaneous comparison
Regression model validated: bootstrap resampling or cross-validation for internal validation; external dataset for prediction models
✏ During Discussion and Interpretation
No causal language used from observational data unless formal causal analysis was performed
OR converted to RR where outcome prevalence > 10% in a case-control study; both reported and discussed
Clinical significance addressed: effect compared to published MCID; no blanket claim that statistical significance implies clinical importance
Non-significant results correctly qualified: if underpowered, p > 0.05 is explicitly described as inconclusive, not as evidence of no effect
Tip: Use It as a Co-Author Review Share this checklist with a co-author or supervisor before submission. A second pair of eyes reviewing specifically for these seven categories — not the clinical content, but the statistical methodology — catches errors that content expertise alone will miss. Many institutions now have biostatistical review services that will complete a checklist like this for free.

Frequently Asked Questions

What is the single most common statistical mistake in published medical research?+
The most pervasive mistake, identified in systematic audits across multiple specialties, is misinterpreting the p-value: specifically, treating p > 0.05 as proof that there is no difference or no effect. Absence of statistical significance is not evidence of absence of an effect — it may simply reflect an underpowered study. A closely related second mistake is applying parametric tests (t-test, ANOVA, Pearson) without first checking normality assumptions. Both errors appear in a substantial minority of published papers across cardiology, surgery, psychiatry, and primary care.
Do I need to check normality before every statistical analysis?+
Yes. Every parametric test — independent t-test, paired t-test, one-way ANOVA, Pearson correlation, linear regression — assumes either that the outcome variable is approximately normally distributed or that residuals are normally distributed. Before running any of these tests, apply the Shapiro-Wilk test (n < 50) or Kolmogorov-Smirnov with Lilliefors correction (n > 50), and inspect a histogram and Q-Q plot. If normality is violated and the sample is small (n < 30), use the non-parametric equivalent: Mann-Whitney U instead of independent t-test, Wilcoxon signed-rank instead of paired t-test, Kruskal-Wallis instead of one-way ANOVA, and Spearman instead of Pearson correlation.
Is p > 0.05 evidence that there is no difference between groups?+
No. This is one of the most dangerous statistical misconceptions in medical research. A p-value above 0.05 means that the observed difference was not statistically distinguishable from sampling error under the null hypothesis — it does not mean the true difference is zero. A non-significant result in a small sample may simply reflect insufficient statistical power. Always examine the 95% confidence interval alongside the p-value: if the CI is wide and includes clinically important effect sizes, you cannot rule out a meaningful effect. The correct phrasing is: "The study was underpowered to detect a clinically important difference" — not "there is no difference."
How many predictors can I include in a logistic regression model?+
The widely cited guideline is the Events Per Variable (EPV) rule: you need at least 10 outcome events for every predictor variable included in the model. For example, if your study has 40 patients who experienced the outcome (e.g., mortality, readmission, complication), you can reliably include at most 4 predictor variables without serious overfitting risk. More recent simulation studies suggest targeting EPV ≥ 20 for stable coefficient estimates, particularly in small samples. Always pre-specify your predictors based on biological plausibility and existing literature — never use stepwise or backward selection driven by p-values.
What is p-hacking and how do statistical reviewers detect it?+
P-hacking (data dredging) occurs when researchers run multiple statistical tests or subgroup analyses and selectively report only those achieving p < 0.05, without correcting for the number of comparisons. Testing a treatment effect across 10 subgroups at α=0.05 without correction gives approximately 40% probability of at least one false positive by chance. Reviewers detect it by looking for: (1) no pre-specified primary outcome, (2) multiple subgroup analyses without Bonferroni or Benjamini-Hochberg correction, (3) all reported results conveniently just below 0.05, and (4) no trial registration record. Prevention: register your primary analysis plan before collecting data.
What should I do if my Shapiro-Wilk test is significant but my sample size is large?+
With large samples (n > 200), the Shapiro-Wilk test becomes extremely sensitive and will flag even minor, clinically trivial deviations from normality as statistically significant. In this situation, visual assessment takes precedence: inspect the histogram and Q-Q plot. If the histogram appears approximately bell-shaped and the Q-Q points follow the diagonal closely, your data is sufficiently normal for parametric tests — the Shapiro-Wilk result is a false alarm driven by sample size. By contrast, with small samples (n < 30), the Shapiro-Wilk test has low power, so a non-significant result does not guarantee normality; visual assessment and knowledge of the variable's structural distribution remain important.
How should I analyse a 5-point Likert scale in a medical thesis?+
For a single 5-point Likert item (e.g., "Rate your pain: 1=no pain to 5=worst imaginable pain"), use non-parametric tests: Mann-Whitney U for two independent groups, Wilcoxon signed-rank for paired/before-after comparisons, Kruskal-Wallis for three or more groups. For a validated composite Likert scale (e.g., sum of 10 items, total score 10–50), first confirm internal consistency (Cronbach's α ≥ 0.70). If the composite score is approximately normally distributed (check histogram + Shapiro-Wilk), parametric tests become more defensible. Always report Cronbach's alpha in your Methods section — failure to do so is a leading peer-review rejection point for PROM-based studies.
Can statistical errors lead to retraction of a published paper?+
Yes. Statistical errors are among the leading documented causes of post-publication corrections and retractions. The Retraction Watch database records hundreds of retractions citing "errors in statistical analysis," "incorrect statistical test," or "data analysis errors." Common retraction-triggering errors include: applying the wrong test to the data type, reporting standard error instead of standard deviation in Table 1 (which artificially inflates apparent precision), calculation errors in reported statistics, and evidence of p-hacking identified through forensic tools like SPRITE, GRIM, or StatCheck. Preventing these errors before submission is incomparably less costly than managing a correction or retraction.
What effect size should I report alongside my p-value?+
The choice depends on your statistical test. For independent t-test: report Cohen's d (small=0.2, medium=0.5, large=0.8). For one-way ANOVA: report eta squared (η²) or partial eta squared (ηp²); small=0.01, medium=0.06, large=0.14. For chi-square: report Cramér's V or phi (φ) for 2×2 tables. For logistic regression: report the odds ratio (OR) with 95% CI. For Pearson or Spearman correlation: the r or rs value is itself the effect size — classify it: small=0.1–0.3, medium=0.3–0.5, large > 0.5. The effect size tells readers whether a statistically significant finding is also clinically meaningful — the p-value alone cannot answer this question.
What is the difference between statistical significance and clinical significance?+
Statistical significance (p < 0.05) tells you that the observed difference is unlikely to have arisen by chance sampling variation. Clinical significance tells you whether the difference is large enough to matter to patients or clinicians. These two concepts can diverge substantially. A large RCT with n=3,000 patients may detect a statistically significant (p < 0.001) reduction in systolic blood pressure of 2.1 mmHg — but the minimum clinically important difference (MCID) for SBP is generally considered 5–10 mmHg. The result is real but clinically trivial. Conversely, an underpowered small study may miss a genuinely large and clinically important treatment effect (p=0.09). Always report effect sizes, 95% CIs, and compare findings to the published MCID for your primary outcome measure.

Analyse Your Data Without the Mistakes

StatClinic's AI-powered tool guides you through test selection, normality checking, and correct interpretation — all without SPSS.

Try StatClinic Free →