Start with the research question
Define the question, available data and what the design can support before choosing software. Prepare a question, unit of analysis, variables or text corpus and intended outcome. Software performs operations; it does not justify your method.
Quantitative route
Check variable types, units, missing-value codes and errors. Describe the sample before comparing groups. Choose methods according to design, dependence and assumptions. Report effect sizes, uncertainty and limits to generalisation alongside statistical significance.
Qualitative route
Define the corpus and analytical approach. Keep a record of reading, codes, categories and revisions. Use extracts to explain your interpretation while protecting participants. Automated summaries or word counts are not substitutes for thematic analysis. Consider cases that challenge your initial account.
Prepare an analysis review
Bring your question, objectives, course guide, data dictionary, de-identified data, transformation record and analysis scripts or coding decisions. Do not upload identifiable participant information through a public form; first agree on authorised sharing.
Traceability and reporting
A comparison of satisfaction scores needs documented recruitment, measurement and missing data. Interviews need an explanation of how extracts became categories. Match each objective to its evidence and analysis, then distinguish findings from interpretation. Return to the final-project hub or request a scoped review.