Key Statistical Principles

Unit 2- Task 2

Key statistical principles of psychometric testing- an overview by Jean Felton

Describe and explain normal distribution and some of the various ways of expressing scores an bands.

Normal distribution ( often described a a 'bell shaped curve') is a term used to describe data that is distributed symmetrically around the curve. This is sometimes referred to as Gaussian distribution. 'Bell curve' refers to the shape that is created when a line is plotted using the data collected. Many distributions of human performance, attributes and characteristics closely resemble the normal distribution and well constructed tests use large samples to ensure it reflects the population's normal distribution accurately.

Scores must be interpreted and compared to the overall data to understand whether the individuals score was within the average range (lie around the centre value) or significantly different to the average.

The standard deviation (SD) measures the variance (how scores vary from the mean) in a standard way. Allowing us to understand how far an individuals score deviates from the mean. For example, you might want to express a score by saying 'it is one standard deviation above the mean.' This would suggest that the result did not deviate greatly from the mean result. However, if the score sat 3 or 4 standard deviations beyond the mean, it is likely to be an unusual result.