Research Methods & Research Literacy
Research literacy is partly about methods and partly about interpretation. The pages here connect study design, statistics, peer review, evidence synthesis, and efficient reading so that a result can be judged in the context of the question it can actually answer.
Research methods
Practical guides to study design, academic reading, peer review, and methodological tradeoffs.
Qualitative vs. Quantitative Research: What Is the Difference?
How qualitative and quantitative methods answer different questions, with sampling, interpretation, limitations, and mixed-methods integration.
Study-design comparisonCross-Sectional vs. Longitudinal Studies: What Changes When You Add Time?
A cross-sectional study is a snapshot; a longitudinal study follows units over time. Compare examples, repeated surveys, change, and limits on causal conclusions.
Evidence-design comparisonRandomized Trials vs. Observational Studies: Strengths, Limits, and Tradeoffs
Compare randomized trials and observational studies through confounding, selection, adherence, generalizability, ethics, and the question being asked.
Research-process explainerWhat Peer Review Can and Cannot Tell You About a Study
What peer review can assess, why publication does not guarantee correctness, and how transparency, replication, and post-publication scrutiny add context.
Reading workflowHow to Read an Academic Paper Efficiently Without Skimming Past the Important Parts
Ways to read academic papers for a specific purpose, from relevance screening to methods, figures, uncertainty, limitations, and usable notes.
Data interpretationHow Missing Data Can Change a Research Result
Why missing observations affect research, what missing-at-random assumptions mean, and how imputation and sensitivity analyses change interpretation.
Study planningPilot and Feasibility Studies: What They Can Establish Before a Larger Trial
How pilot and feasibility studies investigate recruitment, delivery, measurement, and progression decisions without overclaiming treatment effectiveness.
Measurement explainedReliability vs. Validity: What Makes a Research Measure Useful?
Understand reliability, validity, measurement error, and responsiveness through concrete examples of what research scores can and cannot establish.
ExplainerCohort vs. Case-Control Studies: Where the Comparison Begins
A worked comparison of cohort and case-control designs, including retrospective cohorts, control selection, and why an odds ratio is not automatically a risk ratio.
ExplainerWhat Is Mixed Methods Research? More Than Numbers Plus Interviews
How mixed methods research integrates quantitative and qualitative evidence, with examples of explanatory, exploratory, and convergent designs.
ExplainerPreregistration vs. Registered Reports: What Is Committed Before the Results?
How a dated research plan differs from journal review before results, what in-principle acceptance means, and why transparent changes can still be appropriate.
ExplainerSample Size, Statistical Power, and Precision: How They Differ
Why a study can be large without being decisive, how power differs from precision, and why more observations do not fix systematic bias.
ExplainerDoes a Low Survey Response Rate Mean the Results Are Biased?
Why response rate alone cannot establish survey bias, how nonresponse affects estimates, and what weighting can and cannot resolve.
Worked comparisonRandom Sampling vs. Random Assignment: Two Different Uses of Chance
A classroom example separates who enters a study from which intervention they receive.
Practical guideAllocation Concealment vs. Blinding: Before and After Group Assignment
Why hiding the next assignment differs from keeping participants or assessors unaware of assigned treatment.
Focused explainerWhat Is a Cluster-Randomized Trial? Why 1,000 People May Not Be 1,000 Independent Units
A school-based example explains group assignment, within-group similarity, and why the analysis needs to account for clustering.
Worked comparisonCrossover Trials: When Each Person Tries Both Options
How treatment order, washout, and lasting effects affect a study in which participants receive more than one intervention.
Worked comparisonStandard Deviation vs. Standard Error: Variation in People or in an Estimate?
A simple calculation explains why a small standard error does not mean participants had similar outcomes.
Examples in practiceWhat Is a Sensitivity Analysis? Testing How Much a Result Depends on Choices
A worked example distinguishes analytical robustness from proof that every assumption is correct.
Concept comparisonReplication vs. Reproducibility: Same Data, New Data, or a Different Question?
Separate rerunning an analysis from collecting new evidence, while recognizing that fields use the terms differently.
Research literacy
Tools for reading studies, interpreting evidence, and understanding what research can actually support.
Correlation vs. Causation in Health Research: A Practical Guide
How correlation differs from causation, with confounding, reverse causation, statistical adjustment, randomization, and health-research examples.
Information-path analysisHow Research Changes From Paper to Headline to Social Media
How studies change as they become press releases, headlines, and social posts, and which missing details can change what a health claim means.
Reading workflowHow to Read a Health Study Without Overinterpreting It
How to interpret a health study's question, design, participants, outcomes, effect sizes, missing data, limitations, and place in the wider evidence.
Statistics explainerWhat Does Statistical Significance Mean? P-Values and “No Significant Difference”
Statistical significance means a test met a threshold under a model. It does not measure importance, and no significant difference does not prove equality.
Methods explainerSystematic Reviews and Meta-Analyses: What They Can and Cannot Tell You
How systematic reviews and meta-analyses differ, with study quality, heterogeneity, publication bias, forest plots, and certainty of evidence.
Evidence perspectiveWhy One Study Rarely Settles a Health Question
Why health evidence develops across studies, with sampling, design differences, replication, synthesis, and reasons recommendations may change.
Numbers explainedAbsolute vs. Relative Risk: How the Same Result Can Sound Very Different
A worked guide to absolute risk, relative risk, percentage points, and number needed to treat, including baseline risk and time-frame limitations.
Publication explainedPreprints, Corrections, and Retractions: How to Read a Research Paper's Status
What preprint status, published corrections, expressions of concern, and retractions mean when evaluating and citing a scientific claim.
ExplainerConfidence Interval vs. Prediction Interval: Which Uncertainty Is Being Shown?
The difference between uncertainty about an average and uncertainty about a future observation, explained with a worked waiting-time example.
ExplainerPublication Bias and Selective Reporting: When the Missing Studies Matter
How missing studies and missing outcomes can distort an evidence summary, why a large literature is not automatically complete, and what reviewers can examine.
ExplainerStatistical vs. Clinical Significance: How Much Difference Does a Result Make?
Why a statistically significant result may be too small to matter in practice, and why an uncertain result can still include meaningful effects.
ExplainerSurrogate Endpoints vs. Health Outcomes: What Did the Study Actually Improve?
Why a change in a biomarker does not always establish a health benefit, and what makes a surrogate endpoint useful in a particular setting.
ExplainerInternal vs. External Validity: Does a Study Apply Beyond Its Sample?
Internal validity concerns the study’s inference. External validity concerns its relevance elsewhere. Examples explain randomization, target populations, and context.
Concept comparisonNo Evidence of an Effect vs. Evidence of No Meaningful Effect
Why an inconclusive result does not establish equivalence, and how an uncertainty interval changes what can be said.
Focused explainerWhy Significant in One Group Does Not Mean Different Between Groups
A worked example explains the common mistake in subgroup claims and the role of an interaction comparison.
Concept comparisonMean vs. Median in Health Research: Which Average Are You Reading?
Use a five-person example to see why averages differ and why neither one describes everyone's experience.
Worked comparisonA Plausible Mechanism Is Not Proof of a Health Benefit
Why explaining how something could work differs from showing that it improves an outcome people care about.
Focused explainerWhat Does Low-Certainty Evidence Mean?
Understand certainty ratings, why they differ by outcome, and why low certainty does not mean a treatment is definitely ineffective.
Practical guideComposite Endpoints: When One Study Outcome Combines Several Events
How a combined endpoint can be driven by its commonest component, with a worked example and questions for interpreting the result.
Examples in practiceWhy Feeling Better After a Change Does Not Prove the Change Caused It
An explanation of timing, natural variation, and regression toward the mean that takes personal experience seriously without overclaiming causation.