Describing Sets of Observations and Distributions [Intro. 2]

Tukey (1977) has argued that data analysis is like detective work. The investigator examines a set of observations and tries to describe and explain how those observations were generated. From previous experience and knowledge learnt from study, the detective, like the investigator, has from the outset some ideas about what might have occurred. However, there is always the great danger of drawing false conclusions. Careful data analysis, like thorough detective work must be undertaken before an explanation is formally advanced. In data analysis both descriptive and inferential statistics provide the procedures for sound conclusions to be drawn. A detective investigating a robbery uses the best tools as well as an understanding of how robberies are carried out. There are fingerprints everywhere, but the burglar may have left fingerprints at a window or on a drawer where the jewels were kept. Fingerprint powder can be used to detect the fingerprints, but it must be used in appropriate places. Data analysis is very similar. The expedition involves hunting and not merely shooting at what is seen, or line fishing which involves waiting for the fish to bite. Read more ...

Technorati Tags: introduction to statistics, statistics, observations, distribution, central tendency