Four investigators, Iu / 2, / 3, /4, were asked to collect information regarding the proportion of school children in Calcutta, India, between
the ages of 10 and 15 who take milk regularly. Four random samples
of 250 school children each of age group 10-15 were prepared and each
investigator was required to interview students included in his sample.
The following results were obtained.
h h h
Number taking milk 65 38 61 74
Number not taking milk 185 2 1 2 189 176
Should the difference in the proportions observed by different workers
be ascribed to pure chance, or is there any chance to suspect bias or
inaccuracy in the results submitted by any investigator?