Designing a Study You Can Defend
Shared research decisions need not weaken scientific accountability. Learn how to allocate authority, document decisions, manage methodological disagreements, and preserve research integrity in participatory and co-produced studies.
Read Guide →
Designing a Study You Can Defend
Research partners may have strong preferences about what a study should find. Learn how co-produced research can preserve scientific independence through transparent governance, prespecified analyses, conflict-of-interest management, and accurate reporting.
Read Guide →
Designing a Study You Can Defend
Disagreements are a foreseeable part of participatory research. Learn how to establish fair conflict-resolution procedures, distinguish methodological disputes from governance problems, and resolve disagreements without undermining partnership trust, participant rights, or scientific integrity.
Read Guide →
Designing a Study You Can Defend
Community research partners can help identify what information a study should collect, which questions may be inappropriate, and what data are missing from conventional instruments. Learn how to share these decisions while maintaining scientific validity, privacy, and clear data governance.
Read Guide →
Designing a Study You Can Defend
Community research partners can provide contextual knowledge that strengthens the interpretation of findings. Learn when collaborative interpretation is appropriate, how to distinguish evidence from stakeholder explanations, and how to manage disagreements without compromising scientific integrity.
Read Guide →
Designing a Study You Can Defend
Stakeholder participation can appear meaningful while offering little real influence. Learn how to recognize and prevent tokenism through early involvement, clear decision rights, inclusive governance, adequate resources, and transparent reporting of stakeholder contributions.
Read Guide →
Designing a Study You Can Defend
You should usually know how your data will answer your research questions before collecting it. Planning the analysis early can expose design problems, clarify what data you actually need, and reduce data-driven analytical decisions.
Read Guide →
Designing a Study You Can Defend
A useful analysis plan does more than name a statistical test. Before data collection, it should connect each research question to the data, comparisons, analytical methods, assumptions, and decisions needed to answer it.
Read Guide →
Designing a Study You Can Defend
An analysis is appropriate only when it answers the question your study actually asks using evidence your design can legitimately provide. Learn how to test that alignment before collecting data.
Read Guide →
Designing a Study You Can Defend
Every testable hypothesis should have a corresponding analytical path, but the relationship is not always one hypothesis to one statistical test. The analysis must evaluate the specific claim the hypothesis makes.
Read Guide →
Designing a Study You Can Defend
Your research question and inferential goal should drive the study design, not a favorite statistical test. Analysis still belongs in the design process because thinking ahead about statistics can reveal what data the study must collect.
Read Guide →
Designing a Study You Can Defend
You should usually determine your main analytical approach before collecting data, and often the primary statistical method as well. But good planning does not require pretending that every analytical decision can be made before you see the data.
Read Guide →
Designing a Study You Can Defend
Collecting data without knowing how to analyze them does not automatically ruin a study, but it can reveal serious mismatches among the research question, design, measurements, and analysis. The solution is to diagnose the problem before choosing a convenient statistical test.
Read Guide →
Designing a Study You Can Defend
An analysis plan does not need to predict every feature of data that do not yet exist. The important distinction is between decisions that should be made before results are known and flexibility that is genuinely required by the data, methodology, or research purpose.
Read Guide →
Designing a Study You Can Defend
A statistician or methodologist can often contribute most before data collection begins, when the research question, design, measurements, sampling, sample size, and analysis can still be changed.
Read Guide →
Designing a Study You Can Defend
A statistical analysis plan should specify enough detail to make primary analyses reproducible and distinguish planned decisions from those influenced by observed results. Learn what to prespecify, which decisions can remain flexible, and how the required detail varies by study design.
Read Guide →
Designing a Study You Can Defend
Primary analyses should generally be planned before recruitment so study design, sample size, measurement, and inference remain aligned. Learn which decisions need early specification, when later refinements may be acceptable, and how to document legitimate changes.
Read Guide →
Designing a Study You Can Defend
Secondary analyses can answer important research questions beyond a study's primary objective. Learn which analyses should be planned before data collection, how much detail they require, and how to distinguish prespecified secondary findings from exploratory discoveries.
Read Guide →
Designing a Study You Can Defend
Missing data can affect statistical precision, bias, and the interpretation of research findings. Learn how to anticipate missingness during study design, select appropriate analytical strategies, collect useful predictors, and plan sensitivity analyses before the data are collected.
Read Guide →
Designing a Study You Can Defend
Clustering and repeated measurements affect the amount of independent information available in a study. Learn why these design features must be considered before sample size calculations, how correlation influences statistical precision, and why counting observations is not the same as counting independent units.
Read Guide →
Designing a Study You Can Defend
An analysis that cannot answer the research question may signal a fundamental design problem. Learn when to redesign a study, which analytical difficulties can be corrected, and how to preserve scientific integrity when the original design is no longer defensible.
Read Guide →
Designing a Study You Can Defend
Qualitative studies benefit from planning analysis before data collection, but the plan should not predetermine what the data must reveal. Good qualitative planning establishes an analytical direction while preserving the flexibility required by the methodology.
Read Guide →
Designing a Study You Can Defend
You should normally identify a defensible qualitative analysis approach before collecting data, but not every analytic decision needs to be fixed. Learn which commitments matter early and where methodological flexibility is appropriate.
Read Guide →
Designing a Study You Can Defend
Changing your qualitative analysis approach after data collection begins can be methodologically defensible, but not every change is equally consequential. Learn when a revision is justified, what evidence to assess, and how to preserve the integrity of your study.
Read Guide →
Designing a Study You Can Defend
A qualitative codebook can be developed before data collection, but not every analysis approach requires one. Learn when advance coding categories are useful, what a preliminary codebook should contain, and how to avoid restricting unexpected findings.
Read Guide →