126 COST OF LIVING IN THE UNITED STATES 
striking than the differences between them. There is no 
reason to suppose that for its particular coverage each is 
not as nearly correct as possible. 
The Conference Board has found in its experience with a 
number of local investigations of the cost of living, made 
by field agents, that procuring information in this way 
has certain potential dangers. For example, the agent 
arrives at a time when the merchandise manager is busy. He 
asks that the schedule be left; or he hurriedly disposes of 
the agent to buyers or salespeople who, in turn, are busy; 
or he refuses to have the schedule filled out at all. The agent 
may or may not appraise the value of the goods priced and 
its comparability with quotations previously secured. These 
circumstances may result in exactly the same bias, if there 
be any, as in the questionnaire, through inaccurate or in- 
complete quotations which the agent may feel called on to 
adjust to suit the requirements. Such an experience, while 
not necessarily universal, may be frequent. Personal inquiry 
insures nothing regarding the maintenance of a comparable 
standard of merchandise, and leaves much to the discretion 
of the agent. Where, as happens in the Bureau of Labor 
Statistics series, the same agent does not visit the same 
community on two successive dates, there is, of course, 
opportunity for a considerable fluctuation in standard.! 
While collecting data by mail is far from infallible, it has 
the advantage of permitting the merchant to fill in the 
required information at his leisure, and it puts upon him 
the responsibility for seeing that standards of comparability 
are maintained. Its greatest advantage is that it permits 
the collection of data at more frequent intervals, over a much 
wider area, than is possible where special agents have to be 
employed.” A monthly index of any extensiveness could 
hardly be attempted except through questionnaires. Al- 
though the cost of living index of the Bureau of Labor Sta- 
tistics is based on only 32 cities nearly half of the information 
is collected by questionnaire.’ Whether, therefore, question- 
1See, however, Carr, Journal of the American Statistical Association, December, 
1924, op. cit., for different interpretation. 
* Except for local surveys by a local organization. 
3 The various monthly indexes of the Bureau of Labor Statistics are all made up 
from data collected by questionnaire.