ABSTRACT
Iftheanswerstakeanumericalform(intechnicallanguage,ifone isdealingwithavariableratherthananattributequestion),onecan gofurther.Q.S(b),onthesameschedule,askedrespondents,who
ANALYSIS, INTERPRETATION AND PRESENTATION 441
said they had been to the pictures in the last seven days, how often they had gone during this period. The 'Don't know' category apart, the answers would be in numerical form. Again the first step would be to tabulate how many people said 'once', 'twice' and so forth, in fact to make a frequency distribution. The frequencies could then be converted to proportions or percentages as above, but one might also want some measure of average, of what is typical of this sample of respondents. Several kinds of average are available and the researcher must decide which is most suitable to his purpose. Once an average has been calculated, the question arises how representative a figure it is-that is, how closely the answers are bunched around it. Are most of them very close to it or is there a wide range of variation? This calls for one of the several measures of dispersion, and the choice between them again demands judgement.
