extreme statistics

Terms from Statistics for HCI: Making Sense of Quantitative Data

Extreme measures such as maximum and minimum are affected greatly by a small number of values. One consequence of this is that they have relatively complex mathematics. More importantly, they are not robust, that is they tend to vary a lot and if you do the same study or experiment again you are likely to get very different results. In contrast, measures such as the mean (μ), median or quartiles tend to be far more stable.

Defined on page 93

Used on page 93