The remaining sections of this report, their contents and
interrelationships, are described below:
Background, Historical Trends in Article Production, Methods for Counting
Publications and Citations, and Attributing Counts to Institutions and Fields –
presents a brief discussion of the importance and limitations of publication
data as a measure of scientific research, and the concern over an apparent
change in the growth trend of publications that occurred in the early 1990s.
To better quantify and understand possible reasons for this change, this study
examines how trends vary in various academic fields and how institutional
characteristics may influence article production. The section explains that
analyses were conducted using data on the top 200 U.S. academic institutions
with the largest R&D expenditures from 1988 to 2001. A listing of those
institutions can be found in appendix A. The section contains background
information on different methods for counting and attributing publications and
citations, and mapping those counts to fields.
Approach: Scope, Data Sources, and Analysis Methods – provides an overview of
the scope of work, a brief description of the types of variables that were
included in the database, and a general description of analysis methods. Appendices provide important supplemental information to this section. Appendix B contains a more detailed description of the study's methodology. Appendix C contains detailed information about the variables in the database.
Publications, Citations, and Resources – provides graphical displays of trends
in publications, citations, and resources of the top 200 R&D performing
academic institutions, displayed at both the institution and field levels.
Inflation Adjustment and Lagging – describes transformations of variables used
in the regression analyses, including deflation to constant dollars and lagging
(to account for the time between funds receipt, conduct of research,
publication of research finding, and citations to those publications).
of Variability: Between and Within Institution Components – discusses the
finding that almost all of the variability in publication and citation counts
and in resources used for research occurs between institutions rather than
across years within institutions.
Associated with Institution-Level Publication Counts – presents models
developed for fractional and whole publication counts in the expanding journal
set at the institution-year level. From among the many variables considered
for these models, three independent variables were retained: total academic
R&D expenditures, postdoctoral researchers, and Ph.D. recipients. In
addition to the regression model coefficients, the relationships of these
independent variables to publication counts are presented graphically. A path
analytic model is also developed to fit the total effects on publications of
varying academic R&D expenditures. The analyses in this section are restricted
to publication counts in an expanding data set. Appendix D contains a parallel
analysis for citation counts. Appendix F contains a parallel analysis for the
relative citation index (i.e., the number of citations per publication) and appendix
G for fractional and whole publication counts in the fixed journal set.
Model Fit: Time Trends in Resource Utilization Per Publication Count and Model
Fit by Institutional Characteristics – examines whether the adequacy of the
model fit varies as a function of time or institutional characteristics (i.e.,
Carnegie classification, public/private control, patenting activity, amount of
collaboration in publications, and NRC quality ratings). Although no
institutional characteristic was found to affect the fit, there was a clear
linear trend in the ratio of expected to observed publication counts,
indicating that for a given amount of resources the number of fractional and
whole publication counts produced diminished steadily over time.
||Observed and Expected
Publication Trends Within Institutions – presents a refinement to the models of
section 8.0 (which primarily modeled the association of average resources in
the period from 1988 to 2001 to differences in publication counts between
institutions) to allow examination of the effect of changes in resources on
changes in publication counts within institutions. Differences in average
resources available to two different institutions (e.g., one institution with
academic R&D expenditures of $100M and another with $110M) were associated
with larger differences in publications than would be expected if resources
within a particular institution varied by the same amount (i.e., $10M). Within-institution effects are sensitive to the type of funding and type of S&E
postdoctorates, with changes in federally financed R&D funding having a
larger effect within institutions than changes in non-federally financed
academic R&D expenditures.
Associated with Publication Counts at the Field Group Level – presents models
of fractional and whole count publications in the expanding journal set at the
level of five field groups (i.e., biology, life and agricultural sciences; computer
sciences; engineering, math, and physical sciences; medical sciences; and social
sciences and psychology), paralleling the analyses in section 8.
||U.S. Top 200 R&D Performing Academic Institutions – provides information on the top 200
R&D performing academic institutions (as determined by annual average
R&D expenditures from 1988 to 2001), including the campuses and branches
considered to constitute each institution.
Methodology – presents information about the conduct of the study, including
critical issues that were addressed early in the study, steps in database
construction, and the analysis approach.
Descriptions and Database References – provides detailed definitions of the
variables contained in the database and references to the data sources.
of Citation Counts – presents a model of fractional and whole count citations
in the expanding journal set that parallels the analysis in section 8.
||Crosswalk of WebCaspar and ipIQ Classifications – presents a listing of
WebCaspar field definitions and corresponding ipIQ field definitions.
of the Relative Citation Index – presents a model of the relative citation
index (i.e., the ratio of citations to publication counts in the expanding
journal set) that parallels the analysis in section 8.
of Publication Counts in the Fixed Journal Set – presents a model of fractional
and whole publication counts in the fixed journal set that parallels the
analysis in section 8.
||Regression Output for Section 9 presents the output for the regression
analyses referenced in Section 9.
||Regression Output for Section 10 – presents the outpput for the regression
analyses referenced in Section 10.
||Regression Output for Section 11 – presents the output for the regression
analyses referenced in Section 11.