Levels of data, measures of central tendancy, dispersion. and distribution

Solution Preview

DATA ANALYSIS - NUMERICAL SUMMARIES

Levels of data

NOMINAL - This is a simple frequency headcount found in discrete categories (there is only one category a data can go into).

ORDINAL - This is data that can be placed in order, rank or position. The intervals between each rank however are unknown (e.g. how far 1 was from 2).

INTERVAL OR RATIO - This data is on a scale where the intervals are known and are equal. Ratio data has a 'true zero' point. Interval level data can into negative values. This level of data is the most precise.

Measures of central tendency

MODE - This means the 'most frequent'. The mode is used for nominal level data and shows the most popular value:

e.g. 1,2,3,3,4,4,4,5,6 - the mode = 4

MEDIAN - This is the middle value when the scores are placed in rank order. This is most useful for ordinal level data. This tendency is good because it is not distorted by extreme ...