Validating your instrument means checking two things: does it measure what it is supposed to (validity), and does it give consistent results (reliability)? If your questionnaire measures the wrong thing or gives different answers each time, your whole project rests on shaky ground. This guide explains both in plain language and shows how Nigerian students report them.
Validity: Are You Measuring the Right Thing?
- Face validity — at a glance, do the questions look like they measure the topic? Usually judged by your supervisor.
- Content validity — do the questions cover all the important parts of the topic, with nothing missing? Subject experts check this.
- Construct validity — do the questions truly capture the concept (e.g. "motivation")? More common at postgraduate level, sometimes tested statistically.
Reliability: Are the Results Consistent?
Reliability asks whether your instrument gives stable, consistent results. The most common test in Nigerian projects is Cronbach's Alpha, which checks how well the items in a questionnaire hang together. A value of 0.70 or above is generally accepted as reliable.
When You Need Full Validation
- You designed your own questionnaire (most Nigerian projects).
- Your study uses Likert-scale items measuring attitudes or perceptions.
- Your department requires reported validity and reliability (most do).
When It Is Lighter
- You use an already-validated, published instrument — you cite its existing validity and may re-test reliability on your sample.
- Your study is purely qualitative — validity is handled differently (e.g. trustworthiness, member checking), not Cronbach’s Alpha.
Nigerian Project Example
You design a 20-item questionnaire on study habits. Your supervisor and two lecturers review it for face and content validity. You then pilot it with 20 students from a similar class, run Cronbach's Alpha in SPSS, and report a value of 0.81 — above the 0.70 threshold — before using it for the main study.
Undergraduate vs Postgraduate
Undergraduates usually report face and content validity (expert review) plus a reliability coefficient. Postgraduates are often expected to address construct validity too, sometimes with factor analysis, and to discuss validity threats more thoroughly.
Common Mistakes
- Claiming validity without saying who reviewed the instrument or how.
- Reporting a reliability value with no pilot study behind it.
- Inventing or rounding up a Cronbach’s Alpha to clear 0.70 — this is misconduct.
- Confusing validity (measuring the right thing) with reliability (measuring consistently).
Validity and reliability belong in your methodology — see how to write Chapter Three and the full research methodology guide. Project Lab can help you structure clear, testable questionnaire items.
Frequently Asked Questions
What is the difference between validity and reliability?
Validity is whether you measure the right thing; reliability is whether you measure it consistently. An instrument can be reliable (consistent) but still invalid (measuring the wrong thing) — you need both.
What Cronbach’s Alpha value is acceptable?
0.70 or above is generally accepted as reliable in Nigerian projects. If yours is lower, report it honestly and discuss likely reasons — never alter the figure.