This chapter focuses on how tests are devised and used, and how scores on tests may be used to help to draw inferences about personality, intelligence and so on. It considers how correlations can be used when analysing responses to items. The chapter discusses the humble correlation coefficient and shows how it can be used and abused. Correlations are statistics which show the extent to which scores on one variable are proportional to scores on a second variable. The chapter considers amount of new research into how best to quantify the relations between variables. Psychometricians correlate many different types of data. Two continuous variables, such as annual income and years spent in the education system, may be plotted against each other showing each person's data on the two variables as a dot. It is sometimes necessary to calculate the correlation between one dichotomous variable and one continuous variable.