Suchergebnisse

The 1966 meeting of the Amer Assoc for PO Res was studied as an occasion of sci'fic information exchange. Questionnaire's on information-gathering were completed by 186 attendants (78% response). Analysis showed that amount of session attendance, choice of sessions attended, & assessment of information available at the meeting were only weakly predictable on the basis of personal attributes such as level of training, field, seniority, etc. Conversely, information-gathering & -disseminating behavior away from the meeting (eg, N of journal subscriptions, N of technical reports received or sent, N of journal articles published) was highly r'ed with the attributes of commercial/noncommercial affiliation, Sch where highest degree was earned, rank, amount of res activity, level of highest degree, etc. In particular, the strongest predictor of res productivity (defined as written output) is the res'er's level of information acquisition, & an 11-predictor analysis of productivity yielded a multiple r of.70. AA.

10 demographic & econ variables were tested as predicators of library circulation in 2702 US communities. These variables included M educ, F educ, % of white residents, buying income, retail sales, % of Ur'tion, fam size, % of TV saturation, & size of pop served by the library. F educ, income & pop are the best predicators of library circulation, in that order. The multiple r of all predicators with library circulation was .39. The reliability of all is was tested by splitting the sample into random halves & computing is for each half separately. When adult fiction, adult nonfiction, juvenile fiction, & juvenile nonfiction are taken separately, the 10 independent variables predict adult circulation better than juvenile circulation (multiple R's of .44 & .27). In addition, a metropolitan configuration of variables (ie, Urb'tion, pop, TV saturation, income, sales) is more strongly r'ed with nonfiction circulation than with fiction circulation. Library type (eg, city, county) & region (eg, Northeast, Pacific) were included as qualitative predicators of library circulation, & each appears to possess some predictive power when all other variables are held constant. Of methodological interest in this study are: (1) tests of linearity, & (2) the extensive use of eta as a supplementary r'al statistic in the case of non-liner relationships. AA.

When households are the smallest presampled units & random.

Cluster analysis based on the co-occurrence of key words in letters to national magazines yields qualitative indicators of the public's concerns and breaks new ground in content (and historical) research.