Eye strain for office workers

Use this page to show that the marginal probabilities are sums of joint probabilities, such as

Click the formulae for the marginal probabilities to stack the bars.

(The values are actually sample proportions rather than probabilities but the diagram illustrates the concept.)

The diagram is based on a real data set which classifies 295 office workers by their type of work and whether they have symptoms of eye strain.

Data from 295 office workers
Type of work No eye strain Eye strain
VDU data entry 42 11
General VDU use 79 30
Full-time typing 64 14
Standard clerical work 52 3

The sample proportions are treated as joint probabilities in the diagram.