NOTE ON METHOD 139

commissioned to review the literature on the use of the public
library in the United States “today”—adult, young people, and
children; circulation, reference, and other services. The range of
relevance included not only the static description of library use
but also, for example, information on the effectiveness of the
library in influencing reading behavior. It did not include non-
empirical reports on library use, such as exhortations on what the 5C/'-A
public library should do or reports on related aspects of library s
administration and support. Nor were such areas as readability or
reading interests (as distinct from actual reading behavior) con-
sidered appropriate topics for inclusion in this report.

Occasionally in the prosecution of such an analysis of the litera-
ture the area to be covered is redefined in the course of the in-
vestigation as the conception of the field changes or is clarified
or when the literature produces a set of data not originally antici-
~pated. Thus, a certain flexibility in definition must be maintained
—up to a point, at which a literature-searching commitment in
terms of a standard conception of the job must be made. In this
case, for example, the marginal material dealing with people’s atti-
tudes toward library service was added as the literature turned
up a few pieces of data on the subject. On the other hand, only
a minimum amount of material was collected on book reading in b;
general or on book reading from other sources than the pubhc
library or on the general use of the communication media; since
such data was used only for comparative purposes, they werey
secured from the most reliable available sources.

Selection of Studies.—Once the area of interest was determined,
the next step was the selection of appropriate studies for docu-
mentation. Here again, it is only the marginal contributions which
are problematic. There can be no question, of course, about the
inclusion of any empirical study which is more or less directly
concerned with the matter at hand. Nor can there be any ques-
tion about the exclusion of studies with no pretense of sound pro-
cedures for collecting or analyzing evidence. In this connection\
the distinction made for this field between glad tidings, testimony, |
and research is relevant;! items falling into the first two categories '
Beals, 1942.