Page:American Journal of Sociology Volume 15.djvu/633

 CHANGES IN CENSUS METHODS FOR THE CENSUS OF 1910

E. DANA DURAND Director of the Census

It would be quite impossible within any reasonable limits of time to give a general description of the methods which will be employed in taking the coming census. So far as those methods are the same as have hitherto been employed, a description of them would be of little interest to the members of this association, most of whom are already thoroughly familiar with past cen- suses. This paper is, therefore, confined substantially to the differences between the methods which are being or are to be employed in the present census, and those of past censuses.

We believe, of course, that the changes which are to be made will prove advantageous; but one can hold this opinion without in any way disparaging the work of prior censuses. In part the changes proposed are necessary adaptations to actual changes in conditions of our national life. Many of the changes, moreover, are based on recommendations of former Census Directors or of other men whose experience at prior censuses has taught them where the defects lay.

Those changes in census methods which may have a bearing upon the scientific value of the statistics may be grouped under three main heads ; namely, those which relate ( i ) to methods of selecting those employees who collect the statistics; (2) to the scope of the inquiries and the forms of schedules; and (3) to methods of analysis and interpretation. Only as to the first two are plans sufficiently developed to permit any statement of value.

EXAMINATION OF CANIDATES FOR APPOINTMENT

The value of census work depends primarily upon the in- telligence, industry, and integrity of those who collect the sta- tistics in the field. It is, at best, exceedingly difficult to secure competent persons to do the census field-work, because of the very limited duration of the employment which can be offered.

619