Descriptive statistics summarise observed data. Inferential statistics use data and a model to draw conclusions beyond the observed sample, with uncertainty and assumptions. A descriptive percentage does not automatically provide a reliable estimate for a wider population.
Describe the data you actually have
State the number of observations, the variables and the relevant population or setting. Summaries may include frequencies, averages or measures of spread. Choose a summary appropriate to the variable and question. A mean can be useful for some numerical data but is not a meaningful calculation for every kind of category.
Explain the inferential claim
Identify the wider quantity or relationship you want to learn about and the role of the sample. Discuss the selection process and model assumptions. A statistical procedure cannot repair every problem in how the data were collected. Keep the conclusion within the scope justified by the design rather than generalising to everyone from a convenient sample.
Communicate uncertainty
An estimate, confidence interval and hypothesis-test result serve different purposes. Use the interpretation taught by your course and avoid turning a result into a claim of absolute certainty. Statistical significance is not automatically practical importance. Consider the size, context and limitations of an effect alongside the numerical output.
Use frequency terms precisely
In population research, prevalence concerns existing cases or a condition’s presence at a specified time or period; incidence concerns new occurrences over time, with the denominator and measure specified. These are different summaries. This is a terminology distinction for interpreting research, not a way to diagnose an individual. Check the study’s own definition and time frame.
Link outputs to the original question
Explain what each table or calculation contributes. Do not include every available software output without interpretation. Identify missing data, questionable assumptions and relevant limitations. A clear report distinguishes an observed pattern from an inference and explains why the evidence supports that level of conclusion.
Sources and further reading
Related study guides
- Computer-Assisted Personal Interviewing: What CAPI Means
- Types of Research and Study Designs: Choose by the Question
- Convenience Sampling: Meaning, Examples and Limitations
For general study guidance, visit the knowledgebase. Follow your own course requirements when applying this article to assessed work.