NUMBAT OER - Open Educational Resources

3. Analysis of quantitative data - descriptive- and inferential- statistics

Most science is ultimately quantitative, and the hypothesis-led approach demands a quantitative approach both to developing the hypothesis and its predictions, and testing the predictions. Descriptive statistics provide the tools to inspect data, from the simple calculation of averages and measures of variability through to the multivariate techniques used to make sense of very large and complex datasets.

Inferential statistics, on the other hand, are designed to test for patterns in data. Estimation techniques are used to quantify variability, for instance through the construction of confidence intervals. Statistical-hypothesis testing procedures are, as their name suggests, used to test particular types of hypotheses in relation to predictions.