Getting a number from software is not the same as knowing what it means. In a statistics course, the explanation around a result often carries as much weight as the calculation. A useful starting point is to reconnect the output with the question you were trying to answer.
State the question and the data
Identify the outcome, the groups or variables involved, and the population you want to discuss. Then say what data you actually have. A sample collected in one context does not automatically describe every possible population.
Before selecting a method, consider the type of variable and the structure of the observations. Independent groups, paired observations, and repeated measurements are different situations. Bring this decision to tutoring if you are uncertain; choosing a test by its name alone can conceal a mismatch.
Read the estimate before the significance label
Look at the estimated difference, association, or other quantity of interest. Ask about its direction, size, and units. Then consider the uncertainty around it. A label such as ‘significant’ does not tell the whole story and does not by itself establish practical importance.
When interpreting a confidence interval, use the interpretation taught in your course and be precise about the procedure involved. Avoid turning a frequentist interval into a claim that a fixed population value has a stated probability of lying inside that particular realised interval.
Keep causal language under control
An observed association is not automatically a causal effect. Consider how the data were collected and whether the study design supports a causal conclusion. Potential confounding and selection processes may matter more than another decimal place in the output.
A tutor can help you separate what you observed, what the model estimates, and what you are justified in concluding. Working through a fictional example makes it easier to practise the distinction without exposing real personal data.
Write the conclusion for a reader
Use a short structure: answer the question, give the relevant estimate and uncertainty, and mention the main limitation. Explain the result in context rather than pasting an unexplained software table. Keep the wording proportional to the evidence.
For your session, bring the research question, a description of the variables, and a small anonymised or synthetic example. You remain responsible for the analysis you submit. The aim of tutoring is to understand the method and its limits well enough to explain your own decisions.
Keep reading
AP maths: moving from familiar examples to new problems · A more useful way to study calculus