Analysis problems often begin before any test or code is run. A readiness review protects the connection between the research questions, collected evidence, coding decisions, and claims the final report can support.
1. Start with the question and analysis plan
List each research question or hypothesis and identify the exact evidence needed to answer it. For quantitative work, map constructs, variables, levels of measurement, and planned tests. For qualitative work, clarify the analytical approach, role of theory, unit of meaning, and expected audit trail.
- Every question has an evidence source
- The proposed technique fits the design
- Decision rules are written before results are interpreted
2. Protect data meaning
A clean spreadsheet can still be analytically wrong. Check codebooks, value labels, reverse-scored items, units, date formats, skip patterns, derived variables, duplicate records, and the relationship among files.
- Raw data is preserved separately
- Coding rules are documented
- Personally identifying fields are removed where possible
3. Review quality and missingness
Quantitative data requires checks for missing values, impossible values, outliers, distributional assumptions, scale construction, and sample limitations. Qualitative material requires complete transcripts, anonymisation, contextual notes, and a transparent record of exclusions or corrections.
- Corrections never overwrite the raw source
- Missingness or exclusions are reported
- Quality problems are distinguished from inconvenient results
4. Keep interpretation within the design
Statistical significance does not establish importance, causation, or generalisability by itself. A qualitative theme is not credible merely because it sounds persuasive. Interpret effect size, uncertainty, contrary evidence, context, researcher decisions, and limitations.
- Claims match the sampling and design
- Tables or themes answer the research questions
- Alternative explanations and limitations remain visible
5. Prepare reproducible outputs
Save syntax, scripts, project files, coding frameworks, output files, version notes, and decision logs. The final thesis should make the analysis understandable without requiring the reader to reverse-engineer undocumented steps.
- Output names and versions are clear
- Reported values can be traced to analysis output
- The discussion distinguishes results from interpretation
Apply the checklist to your project.
Download the printable resource, review the related service, or send a structured enquiry with your programme, current stage, exact issue, and deadline.