NOTE ON METHOD 141

used, as in the cases of age or occupation or size of city, and it
accordingly became difficult to “match up” data on the same
problem. Or different bases were used for essentially the same
problem. For example, some studies would report the use of
the public library as a source of books in terms of the percentage
of book readers served by the library, and others would report
such data in terms of the percentage of books supplied by the
public library. There is no one-to-one relationship necessary, of
course, because of the multiple use of sources and the different
rates of use of each source.

On some occasions it became possible to retabulate or recom-
pute published data to maximize their correspondence with simi-
lar data from other studies. For example, some of the material on
the extent to which different (adult) groups within the com-
munity use the public library (as reported in Chapter 2) was re-
analyzed from data presented in the incidence of such groups
within the library clientele. On other occasions, special tabula-
tions were secured from some studies? in order to add significant
data to the total body of information. The diversity of the studies
in almost every field of library use not only makes integration dif-
ficult but also places an additional burden upon the interpreta-
tion of results.

Interpretation of Data.—Drawing generalizations from such
heterogeneous material is hazardous. Caution has been exercised
at many points in the report where the data are especially diverse.

There are two kinds of generalizations about the use of libraries
which involve different bases of interpretation. In the first place,
there are generalizations about incidence of library use. How
many people “use” the public library? What is the frequency
distribution of different educational groups in the library clien-
tele? In what subject areas do reference questions fall? The an-
swers to such questions obviously depend upon the location of
the studies; the answer for a well-to-do suburb may be quite dif-
ferent from that for an industrial community. In such cases, the
nature of the individual studies was particularly taken into ac-
count and some judgment was exercised in generalizing from
*For example, SRC, 1948, and Field and Peacock, 1948.