Practical guides written for clinicians and researchers not statisticians. From choosing the right test to writing your thesis methods section.
Choosing, running, and interpreting statistical tests
When to use it, adjusted means, homogeneity of regression slopes, how to choose a covariate, a worked medical example, a step-by-step StatClinic walkthrough, and how to interpret and report your results.
When to use it, unequal variances and sample sizes, Levene's test limitations, a worked medical example, a step-by-step StatClinic walkthrough, and how to interpret and report your results.
When to use it, assumptions, Wilks' Lambda vs Pillai's Trace, a worked medical example, a step-by-step StatClinic walkthrough, follow-up tests, and how to interpret and report your results.
When to use it, assumptions, a worked medical example, a step-by-step StatClinic walkthrough, post hoc comparisons with multiple-testing correction, and how to interpret and report your results.
When to use it, assumptions, a worked medical example, a step-by-step StatClinic walkthrough, Dunn's post hoc comparisons, and how to interpret and report your results.
Sphericity, Mauchly's test, the Greenhouse-Geisser correction, a worked longitudinal example, a step-by-step StatClinic walkthrough, and how to interpret and report your results.
Main effects, interaction effects, factorial study designs, assumptions, a worked clinical example, a step-by-step StatClinic walkthrough, and how to interpret and report your results.
What the Wilcoxon signed-rank test is, when to use it, assumptions, a worked clinical example, a step-by-step StatClinic walkthrough, and how to interpret and report your results.
Normality, independence, homogeneity of variance, linearity, outliers, multicollinearity, expected cell counts, and the proportional hazards assumption explained with clinical examples and a practical checking workflow.
Cross-sectional, case-control, cohort, randomized controlled trials, quasi-experimental designs, and systematic reviews explained with clinical examples the essential first step before choosing your variables or statistical test.
Independent, dependent, confounding, continuous, categorical, binary, nominal, ordinal, time-to-event, and repeated measures variables explained with clinical examples the essential first step before choosing a statistical test.
A complete decision framework from study objective to the exact test with a decision table covering 15+ scenarios. Includes parametric vs non-parametric, normality testing, and common mistakes.
The complete guide to choosing between paired and independent t tests with real clinical examples, a decision table, common mistakes to avoid, and an FAQ covering 5 key questions.
The expert guide to understanding what p < 0.05 really means, why researchers misread it, the critical difference between statistical and clinical significance, and how to report p-values correctly in papers and theses.
Learn the number of groups rule, why multiple t tests inflate error rates to 26.5%, and how to choose between t test, one-way ANOVA, and repeated measures ANOVA with 4 clinical scenarios and a decision flowchart.
Master normality testing Shapiro-Wilk vs KS comparison, histogram and Q-Q plot interpretation, SPSS step-by-step guide, 4 worked medical examples, and the parametric vs non-parametric decision framework.
Master the 5-cell rule for expected frequencies, understand when small samples demand Fisher's Exact, and see two fully worked contingency table examples antibiotic trial (n=200) and immunotherapy pilot (n=15).
Master Kaplan-Meier survival curves, censored data, median survival time, and the log-rank test with three oncology examples, common mistakes to avoid, and a complete reporting guide.
The complete guide to binary logistic regression when to use it, how to build and check the model, how to interpret odds ratios, three worked clinical examples, and a full journal reporting checklist.
Master the 2-2 contingency table learn why a 95% sensitive test can have a positive predictive value below 20%, how prevalence changes everything, and how to correctly interpret diagnostic test results in clinical practice.
Master multivariable regression in clinical research confounding variables, adjusted odds ratios, logistic vs linear vs Cox regression, model building strategy, and common mistakes to avoid.
The complete guide to confidence intervals what 95% CI means, formula components, precision and sample size, 4 clinical examples for means/proportions/OR/RR, journal reporting formats, and 6 common mistakes with fixes.
Step-by-step guide to calculating OR from 2x2 tables, interpreting crude vs adjusted OR, reading logistic regression Exp(B) output, 4 worked clinical examples including smoking and lung cancer, and 6 common mistakes to avoid.
Confounder vs mediator vs effect modifier explained, clinical knowledge and literature review, a DAG introduction, a covariate decision flowchart, purposeful vs stepwise selection, overadjustment bias, multicollinearity, and sample size rules with logistic and linear regression examples.
A beginner-friendly guide to Cox proportional hazards regression: hazard vs risk, censoring, hazard ratio explained with simple examples, adjusted vs unadjusted HR, SPSS output tables, confidence intervals, the proportional hazards assumption, and a step-by-step medical example.
Every part of a forest plot explained simply study names, squares, confidence intervals, the diamond, weight, the line of no effect, and heterogeneity with two fully labeled example plots, fixed vs random-effects models, and forest plots for OR, RR, MD, SMD, and HR.
Assumptions, mathematical intuition without equations, advantages and disadvantages, how heterogeneity (Iยฒ and Tauยฒ) drives model choice, effects on confidence intervals and pooled estimates, RevMan examples, a comparison table, and a text-form decision flowchart.
Why publication bias occurs, small-study effects, symmetrical vs asymmetrical funnel plots, Egger's test, Begg's test, trim-and-fill, common misconceptions, and practical RevMan, Stata, and R examples for your systematic review.
Coefficients (B), standardized vs unstandardized Beta, confidence intervals, p-values, Rยฒ and adjusted Rยฒ, assumptions, multicollinearity, residual analysis, and SPSS output, explained simply with a complete worked medical example and reporting checklist.
Expert guide to choosing between mean and median: how skewed distributions and outliers force the choice, the outlier demonstration with ICU length of stay, five clinical dataset examples with verdict and phrasing, the link to parametric vs non-parametric tests, a step-by-step decision framework, six common mistakes, and correct journal reporting formats.
Complete expert guide to effect size: why p-value alone is insufficient, Cohen's d formula step by step, small/medium/large interpretation thresholds, four clinical trial examples with full calculations, statistical vs practical significance, ANOVA eta squared and omega squared, effect sizes for non-parametric tests, six common mistakes, and journal reporting standards (APA, CONSORT).
Complete expert guide to ROC curve analysis: AUC interpretation (0.7โ0.8โ0.9 scale), the sensitivity-specificity trade-off, optimal cutoff selection with Youden Index, four clinical laboratory examples (troponin, HbA1c, D-dimer, procalcitonin), good vs poor diagnostic tests, DeLong AUC comparison test, common mistakes, and journal reporting formats.
Expert comparison of Pearson r and Spearman rₛ: formulas, assumptions, 6-step decision framework, four worked clinical examples with full calculations, interpretation of weak/moderate/strong correlation, coefficient of determination, common mistakes, and journal reporting formats.
Expert comparison of RR and OR: definitions, formulas, study design rules, the rare disease assumption, when OR overestimates RR, four clinical worked examples side by side, Zhang & Yu conversion formula, common mistakes, and reporting formats.
Paired categorical outcomes โ before/after, matched case-control โ where chi-square fails. Covers discordant pairs logic, McNemar formula, Yates correction for small n, exact binomial version, and SPSS/R reporting.
Agreement vs chance-corrected agreement โ why raw percentage agreement misleads. Covers kappa formula, weighted kappa for ordinal categories, strength benchmarks (Landis & Koch), and the high-agreementโlow-kappa paradox.
6-model selection table (one-way / two-way ร absolute / consistency ร single / average), ICC reliability thresholds, SEM-to-MDC calculation, and 3 clinical examples from ECG measurement to physiotherapy assessment.
Systematic framework for comparing two measurement methods: bias line, 95% limits of agreement, proportional bias regression test, and 3 clinical scenarios showing when Bland-Altman passes but clinical equivalence still fails.
ANOVA tells you significance exists; post hoc tells you where. Full comparison of Tukey HSD vs Bonferroni with FWER formulas, worked examples, when Scheffรฉ and Games-Howell are needed, and thesis reporting templates.
Why controlling for confounders can be wrong when the variable is an effect modifier. Covers stratification, Mantel-Haenszel pooled OR, interaction term testing in regression, and 4 clinical examples that make the distinction concrete.
Same two variables, two very different answers โ when Pearson r and Spearman ฯ diverge and why. Covers linearity vs monotonicity, partial correlations, 3 clinical datasets with full calculation comparison, and thesis reporting formats.
When small n, severe outliers, or non-normality make t-test unreliable โ Mann-Whitney U as the alternative. Covers rank sums, exact vs asymptotic p, Hodges-Lehmann estimator, ARE efficiency under normality, and effect size r.
P < 0.05 says the effect exists; effect size says whether it matters clinically. Covers Cohen's d, OR, RR, NNT, and r with standard benchmarks, the minimal clinically important difference concept, and 3 trials where significance misled.
Detection toolkit: Z-score, modified Z-score, IQR rule, Grubbs test, Cook's D, and Mahalanobis distance ranked by use case. Includes a 5-step decision protocol โ investigate, verify, decide, transform, report โ with 3 clinical examples.
Normality testing by sample size, ARE efficiency (Mann-Whitney U = 0.955ร t-test power under normality), 7-row test equivalents table, and a 6-step decision flowchart that takes you from raw data to the correct test in minutes.
With 20 tests at ฮฑ = 0.05, the chance of โฅ1 false positive exceeds 64%. Covers FWER table, 4 correction method cards (Bonferroni / Holm / ล idรกk / BH-FDR), worked Benjamini-Hochberg example, and 6 pre-specification strategies.
SPSS, questionnaires, reporting, and thesis methodology
An objective, practical comparison covering ease of learning, supported tests, clinical workflow, AI-assisted analysis, and cost with clear recommendations for students, residents, clinicians, and researchers.
Variable naming, patient IDs, dates, missing values, categorical coding, merged cells, duplicates, and outliers the complete beginner guide to preparing a clinical Excel dataset, with good-vs-bad spreadsheet examples throughout.
Step-by-step guide to coding, cleaning, and analyzing questionnaire data. Covers Likert scale analysis, Cronbach's alpha reliability, Chi-square, Mann-Whitney, and Spearman correlation.
Step-by-step SPSS guide from data entry to output interpretation variable coding, descriptive statistics, normality testing, t-test, ANOVA, Chi-Square, and 6 common thesis mistakes with fixes.
A table-by-table guide to every major SPSS output descriptives, t-test, ANOVA, chi-square, correlation, regression, ROC curve, and Cronbach's alpha with real examples, wrong-vs-right interpretations, and journal-ready reporting.
Master Cronbach's alpha from formula to SPSS output interpretation thresholds, acceptable values by research context, validity vs reliability distinction, and 6 common mistakes with fixes.
Upload your Excel dataset and let AI automatically detect variable types, select the right statistical test, interpret your results in academic language, and generate a downloadable PDF report free, browser-based, no SPSS license needed.
A balanced, educational look at the SPSS learning curve, how guided and AI-assisted platforms simplify analysis and reduce errors, a side-by-side workflow comparison, and the situations where SPSS still remains the better choice.
Comprehensive expert guide to the 7 most dangerous statistical errors in medical research: wrong test selection, ignoring normality assumptions, inadequate sample size, p-value misuse and p-hacking, overfitting regression models, poor questionnaire and Likert scale analysis, and misinterpreting results โ with real clinical scenarios, journal publication consequences, incorrect vs correct comparisons, and a practical pre-submission prevention checklist.
Complete expert guide to reporting statistics in medical papers and theses: APA 7, CONSORT, STROBE, and PRISMA standards; correct formats for p-values, confidence intervals, t-test, ANOVA, chi-square, odds ratio, relative risk, linear and logistic regression, non-parametric tests, and correlations; incorrect vs correct comparison blocks; a master reference table; thesis-specific recommendations; and 10 common reporting mistakes with fixes.
A reporting template, real example, incorrect example, and corrected version for 20 statistical tests from descriptive statistics through Kaplan-Meier, Cox regression, ROC, ICC, Bland-Altman, and Kappa aligned with APA 7, CONSORT, STROBE, and PRISMA, plus a universal checklist and a downloadable template.
12 real peer-review statistics comments medical journals raise most often wrong test, no normality check, sample size justification, missing CIs and effect sizes, regression assumptions, multicollinearity, multiple comparisons, missing data each with why reviewers ask it, how to fix it, and a response-to-reviewers example.
Every section of a SAP explained in plain language primary and secondary outcomes, variable definitions, statistical methods, missing data and outlier strategy, subgroup and sensitivity analyses, and software with a complete sample SAP, common mistakes, and a downloadable template.
Content validity ratio, item-level CVI, scale-level CVI, Delphi panel setup, forward-backward translation, factor analysis essentials, and test-retest reliability โ a complete questionnaire validation workflow for thesis researchers.
MCAR, MAR, and MNAR missingness with Little's MCAR test; complete-case analysis bias quantified; multiple imputation (MICE) step-by-step; LOCF vs BOCF trade-offs; sensitivity analysis; and journal-level reporting standards.
4 determinants of power (ฮฑ, ฮฒ, effect size, n) in one unified framework. Power tables, G*Power input walkthrough, the post-hoc power fallacy, consequences of underpowered studies, and 3 full sample size calculations with formulas.
ITT preserves randomization integrity; per-protocol estimates biological efficacy โ knowing when to report both is essential, especially for non-inferiority trials. Covers modified ITT, MMRM missing data, CACE estimation, and CONSORT Items 13/16/17.
Sample size and power for every study design
Master the Cochran formula for cross-sectional studies. Includes design effect adjustment, finite population correction, non-response inflation, and three fully worked examples.
Master the Kelsey formula for unmatched case-control studies step-by-step with the OR-to-p conversion, three fully worked epidemiology examples (COPD, CKD, C. difficile), OR impact table, and 6 common researcher mistakes with fixes.
Complete step-by-step guide to cohort study sample size: the two-proportion formula, all variables explained, prospective vs retrospective design, four worked epidemiology examples, loss-to-follow-up adjustment, power and confidence level impact, common mistakes, and STROBE-compliant reporting formats.
Jump straight to the guides most relevant to your research.