APPLIED SOCIOLOGYMapping and social indicator research. Such
studies are primarily descriptive, and designed to
provide estimates of the incidence and prevalence
of phenomena that are social in nature. There may
be interest in how these phenomena are distribut-
ed in different social categories (e.g., by ethnic
group affiliation, lifestyles, or social classes) or are
changing over time. For example, corporations
may wish to know how consumption patterns for
various goods are changing over time for different
groups to facilitate the development of marketing
strategies. Federal and state agencies may wish to
understand how the incidence and prevalence of
diseases for different social groups is changing
over time to develop more effective prevention
and treatment strategies. Neighborhood groups
may wish to know how citizen complaints regard-
ing the police, high school graduation, or cervical
cancer rates are distributed in relation to income
characteristics of neighborhoods.
It is often assumed that applied sociology is
less rigorous than basic sociology. Freeman and
Rossi (1984) suggest that it is more appropriate to
argue that the norms governing the conduct of
basic sociological research are universally rigor-
ous, while the norms governing the conduct of
applied sociological research, of which mapping
and social indicator research are but two exam-
ples, have a sliding scale of rigor. For critical
decisions, complex phenomena that are either
difficult to measure or disentangle, or where pre-
cise projections are needed, sophisticated quanti-
tative or qualitative measures, or both, may be
required and sophisticated analytic techniques may
be needed, or may need to be developed. But as
time and budget constraints increase, and the
need for precision decreases, ‘‘quick and dirty’’
measures might be more appropriate. The level of
rigor in applied sociology, in short, is driven by the
needs of the client and the situation as well as the
nature of the problem under investigation.
Modeling social phenomena. The modeling
of social phenomena is an activity common to both
basic and applied sociology. Sociologists of both
persuasions might be interested in modeling the
paths by which adolescents develop adaptive or
maladaptive coping strategies, or the mechanisms
by which social order is maintained in illicit drug
networks. The applied sociologist would need to
go beyond the development of these causal mod-
els. For the adolescents, applied sociologists would
need to understand how various interventions
might increase the development and maintenance
of adaptive strategies. In the case of drug net-
works, they might need to understand the relative
effectiveness of different crime-control strategies
in reducing the capacity of drug networks to pro-
tect themselves.Just as with mapping and social indicator re-
search, while there are norms supporting rigor to
which basic researchers should adhere in the de-
velopment of causal models, there is usually a
sliding scale of rigor in applied research. There is
not a necessary trade-off between doing applied
work and levels of rigor. It usually happens, howev-
er, that applied problems are relatively more com-
plex and tap into concepts that are less easy to
quantify. In trying to capture more of the com-
plexity, precision in the specification of concepts
and elegance of form of the overall model may be
traded off against an understanding of dynamics
that is sufficient to inform decisions.For both basic and applied work it is impera-
tive that the mechanisms by which controllable
and uncontrollable concepts have an impact on
the phenomenon of interest are properly speci-
fied. Specification decisions must be made with
respect to three aspects of the dynamics of situa-
tions (Britt 1997). First, the nature of what con-
cepts are (and what they are not) must be specified
with regard to the contextually specific indicators
that give them meaning. Second, the concepts that
are considered sufficiently important to be includ-
ed in a model (as well as those that are deemed
sufficiently unimportant to be left out) must be
specified. Finally, the nature of relationships that
exist (and do not exist) among the included con-
cepts must be specified. Since these specification
decisions interpenetrate one another (i.e., have
implications for one another) over time, it may be
prudent to think of models as vehicles for summa-
rizing what we think we know about the dynamics
of situations (Britt 1997).The clients of applied researchers, however,
are not interested in how elegant models are, but
in how well the implications of these models assist
them in reducing the uncertainty associated with
decisions that must be made. As time and budget
constraints increase, or researchers become more
confident in their ability to understand which
controllable concepts are having the biggest (or