Resources

Methods notes written for people running studies, not for people selling to them. Every article names its author and cites what it claims.

Systematic reviews

Protocols, search strategies, screening, extraction and risk of bias — the decisions a review is judged on.

How to build a systematic review search strategy
A search is reproducible or it is not a systematic search. Here is how to build one that another researcher could run and get the same results.
2 min read
GRADE: what certainty of evidence means and how it is decided
A pooled estimate says what the evidence shows. Certainty says how much the estimate is likely to change when better evidence arrives — a separate and often more useful question.
2 min read
Risk of bias: choosing a tool and using it properly
Risk of bias is an assessment of whether a study's result can be believed, not a score of how well it was written up. The distinction changes how the assessment is done.
2 min read
How to write a systematic review protocol
A protocol is what separates a systematic review from a literature search performed carefully. Here is what it has to decide, and why each decision has to be made before the results are known.
2 min read

Pooling models, heterogeneity, sensitivity analysis and the question of whether to pool at all.

Publication bias: funnel plots, Egger's test and what they can and cannot show
Funnel plot asymmetry has several explanations and publication bias is only one of them. Reading it as proof of missing studies is a common and consequential error.
2 min read
What is heterogeneity in meta-analysis, and what does I² actually mean?
I² is widely reported and widely misread. It does not measure how different the studies are, and it is not a threshold for whether pooling is allowed.
2 min read
Fixed-effect vs random-effects meta-analysis
The two models answer different questions. Choosing between them by which one gives a narrower interval is choosing the answer before asking the question.
2 min read
When should you perform a meta-analysis — and when should you not?
Software will pool almost anything you give it. The methodological question is whether a single number describing these particular studies means anything at all.
2 min read

Choosing methods that match the design, and reading the output as something other than a verdict.

Survival analysis: Kaplan-Meier, Cox regression and competing risks
Time-to-event methods exist because censoring is information. The commonest errors come from forgetting that, or from ignoring the events that make the event of interest impossible.
2 min read
Missing data: complete-case analysis, imputation and what each assumes
Every method for handling missing data makes an assumption about why it is missing. The assumption is usually untestable, which is exactly why it has to be stated.
2 min read
Sample size and statistical power: what the calculation actually needs
A sample size is not a number you find; it is the consequence of decisions you make. Most calculations fail because one of those decisions was never made explicitly.
2 min read
Confounding, mediation and collider bias: what to adjust for
Which variables belong in a model is a question about causal structure, not about which ones reach significance. Adjusting for the wrong one introduces bias that was not there before.
2 min read
Odds ratios, risk ratios and hazard ratios: which to report
The three are routinely described as interchangeable and are not. Reporting the wrong one, or interpreting it as though it were another, changes what the paper claims.
2 min read
How to choose a statistical test
The test is the last decision, not the first. Almost every wrong test in a medical paper is a symptom of a question that was never stated precisely.
2 min read

What honest reporting looks like in practice, and how to recognise its absence.

How to respond to reviewer comments
Most reviewer objections are about a methodological decision that was never explained. The response that works says what changed, where, and why — including where nothing changed.
2 min read
Reporting guidelines: which one applies, and how to use it properly
A checklist submitted with the manuscript is not the same as a study reported completely. Used at the end, guidelines catch omissions; used at the start, they prevent them.
2 min read
Why a guaranteed publication is a red flag
No one can guarantee peer review. A service that promises it is either not describing peer review, or is describing something you do not want your name on.
2 min read

Observational designs, the biases that live in routinely collected data, and how far the conclusions can honestly go.

Keywords: How Database Indexing Drives Your Paper’s Citations
Search engines index title words automatically. Keywords exist to add what your title missed—mapping your paper into major database index hierarchies (PubMed MeSH, Embase Emtree) so researchers actually find and cite it.
1 min read
Propensity scores: matching, weighting and what they cannot fix
A propensity score balances the variables you measured. It does nothing about the ones you did not, which in observational research are usually the ones that mattered.
2 min read
Immortal time bias, and how to avoid building it into your study
A period during which the outcome could not have happened, assigned to one group by the way the study defines exposure. It reliably makes treatments look protective.
2 min read
What is real-world evidence, and what can it actually answer?
Real-world data is data collected for a purpose other than your study. That single fact explains most of both its value and its limits.
2 min read

Send the scope and we will come back with a written quote. No account needed.