Navigation: »No topics above this level«
Observed datasets can be described by listing every recorded value, but this gives a very confusing and unwieldy picture of the information available. Despite this, careful inspection of sample data values is often an essential and important part of any data analysis exercise. Initial inspection might include an examination of the mix of data, the occurrence of duplicate values, and checking for missing or mis-coded data, zero values and unexpected values (e.g. negatives when none are expected). Subsequently a range of descriptive statistics may be produced, which seek to summarize the data in some way, and is often complemented by simple graphical representations such as line graphs, histograms or scatter diagrams. In the following sections many of the common descriptive statistics are described.