Campus Tech
Student data projects that actually get published
GOSPELTRADER Research Desk · 12 August 2026 · 7 min read
Quick answer
Projects pass when the question is narrow, the instrument is piloted, the sample size is justified in writing and every result can be re-run from the submitted files. Most rejections trace back to one of those four.
We review a large number of student projects each session. The failure patterns are consistent enough to list.
1. Narrow the question until it is boring
'The impact of social media on students' is not a research question; it is a topic area. 'The relationship between daily short-video use and reported study hours among 300-level students in one faculty' is answerable with the resources a student actually has.
2. Pilot the instrument on ten people
Ten pilot responses will expose every ambiguous question, every missing option and every item nobody understands. This single step saves more projects than any statistical technique.
3. Justify the sample size in writing
State the formula, the assumed proportion, the confidence level and the margin of error, then show the arithmetic. A number with no derivation invites the first question at defence.
- • State the population frame
- • Name the formula used
- • Show the parameters and the result
- • Report the achieved response rate honestly
4. Make it re-runnable
Submit the cleaned dataset, the syntax or script and a codebook. A supervisor who can reproduce your table in five minutes will argue with your interpretation, not your competence.
Free tools we publish for this
Our Labs section includes a sample size calculator, a statistical method recommender, a research topic evaluator and a data quality checker — all free with a student account.
Frequently asked questions
Which software should a student use?
Whatever the department supports and can review. SPSS remains standard for many Nigerian faculties; R and Python are stronger for reproducibility if your supervisor accepts them.