140 NOTE ON METHOD

were omitted in this report. Such “studies” have little or no
scientific basis, and the reader has small opportunity to evaluate
the findings. In addition, such reports typically do not provide the
required degree of precision in the presentation of findings.

Not everything in every study was utilized in the final version
of the report, although everything was scrutinized for inclusion.
Some parts of some studies (and even a few complete studies) were
omitted because of basic deficiencies in the collection and analy-
sis of the data. A few studies were excluded on the grounds of
too great specificity. Finally, not much use was made of the an-
nual reports of particular libraries. First, there was the difficulty
of securing an adequate sample of such reports, and, second,
with a few exceptions they yield only broad data of the sort con-
tained in the typical statistical compilations (for example, per
capita circulation, circulation by fiction and nonfiction, and so
forth).

Every effort was made to select all those studies which might
contribute validly and reliably to the central problem of de-
termining the extent and the nature of public library use. A mini-
mum level of “scientific” performance was demanded of the
studies selected for inclusion. Even so, because of the nature of
the field, doubtful material was included at several points; the
compilers erred in the direction of inclusion.

Integration of Data.—Once the studies were chosen, the basic
problems of integration and interpretation began. The central
point about the integration of the research literature is that the
unit is the datum, not the study. Findings from various studies
were put side by side for purposes of summarization, and differ-
ent findings from the same study may be found throughout the
report. In this connection, the integrity of problem areas or of
subject matter was observed rather than the integrity of individ-
ual studies.

Because of the extreme diversity of the studies included in this
critical summary, the problem of integration was not easily
solved. Unfortunately studies have a way of suiting their own
purposes rather than the summarizer’s purpose. For example,
different classifications for the same objective characteristics were