Conover PeopleOrganizePolitical 1984

How People Organize the Political World: A Schematic Model
Author(s): Pamela Johnston Conover and Stanley Feldman

Source: American Journal of Political Science, Vol. 28, No. 1 (Feb., 1984), pp. 95-126
Published by: Midwest Political Science Association
Stable URL: https://www.jstor.org/stable/2110789
Accessed: 18-12-2024 03:52 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Midwest Political Science Association is collaborating with JSTOR to digitize, preserve and
extend access to American Journal of Political Science
This content downloaded from 172.96.140.89 on Wed, 18 Dec 2024 03:52:46 UTC
All use subject to https://about.jstor.org/terms
How People Organize the Political World:
A Schematic Model *
Pamela Johnston Conover, Stanley Feldman, University of Kentucky
A schematic model of political belief systems based on recent research in social and
cognitive psychology is developed. We show that schema theory has the ability to bridge
the gaps among the major competing approaches to the study of mass belief systems as
well as provide direction for new research on the structure and functioning of political
beliefs. An initial test of the theory, done with Q methodology, is reported and the results
provide consistent support for the general predictions of schema theory as applied to
political beliefs.
In the study of belief systems there has recently been a sharp increase
in the number of writings critical of "paradigmatic" understandings, and
there is a growing lack of consensus in the literature about the meaning
of key concepts (Bennett, 1977). To some extent, this confusion has been
generated by research from basically two different perspectives. Some
theorists advocate a "sociological" view that emphasizes the "social"
origins of mass belief systems and focuses on the constraint or structure
among specific issue positions (e.g., Converse, 1964; Nie, Verba, and
Petrocik, 1976). Others offer a "psychological" perspective that stresses
the individualistic origins of belief systems and that concentrates on the
structure between specific issue positions and more general political ideas
(e.g., Lane, 1973; Marcus, Tabb, and Sullivan, 1974; Jackson and Marcus,
1975; and Thomas, 1978). Despite their differences, these two perspectives
have shared a common concern with the question of whether people
think ideologically. Yet, recently even this shared focus has come under
fire. Kinder (1982), for one, has argued that, since the great bulk of
research indicates that most people do not think ideologically, we should
abandon our focus on the ideology question and concentrate instead on
how people actually think about politics.
In this paper we do just that. Specifically we propose a schematic
model of how people organize their beliefs about the political world.
While this model by no means resolves the debate between the sociological
and psychological approaches to political belief systems, it does provide
a basis for integrating certain aspects of the two frameworks. More
*An earlier version of this paper was presented at the 1982 Annual Meeting of the
American Political Science Association, the Denver Hilton Hotel, September 2-5, 1982.
The order of the authors' names is alphabetical; we each share equally in any credit that
might be due. We would like to thank Bruce Campbell and several anonymous referees for
their helpful comments on an earlier version of this paper.
96 Pamela Johnston Conover, Stanley Feldman
important, it also leads us to consider phenomena generally not dealt
with by the traditional literature on mass belief systems. Once we have
outlined our approach we proceed to specify a means of testing our
theory that draws heavily on Q methodology. Finally, we review our
initial efforts to test some of the aspects of the model.
Schema Theory
In recent years a substantial metatheoretical shift has taken place in

the study of social cognition. The mainstream of social psychology has
moved away from cognitive consistency theories and their assumption
that people seek cognitive consistency and hence are "rationalizing,
motivating, face-saving, and justifying" (Taylor, 1981, p.192). From this
heavily motivational view of people, a different perspective has gradually
evolved: that of people as "cognitive misers" who have a limited capacity
for dealing with information, and thus must use cues and previously
stored knowledge to reach judgments and decisions as accurately and
efficiently as possible. This is not to say that people are simply "cold"
information processors who always handle information in a rational,
orderly fashion. On the contrary, this view implies that, because people
have a limited capacity, they are often forced to make decisions and
judgments on the basis of scanty data that may be "haphazardly
combined and strongly influenced by preconceptions" (Taylor, 1981, p.
194).
This paradigmatic shift in perspectives has led psychologists to focus
more on how knowledge is stored and how such stored information
subsequently influences the perceptual process. In such efforts, the
concept of a "schema" has played a central role (Neisser, 1976; Markus,
1977; Hastie, 1981; and Taylor and Crocker, 1981). A schema may be
defined as a cognitive structure of "organized prior knowledge, abstracted
from experience with specific instances" that guides "the processing of
new information and the retrieval of stored information" (Fiske and
Linville, 1980, p. 543). For example, a schema of the role of "candidate"
might include very general beliefs about the goals of candidates along
with more specific information about the particular activities that candi-
dates engage in to get elected. For our purposes, we assume that schemas
are content specific (Taylor and Crocker, 1981), and that different types
of schemas may therefore be identified according to the nature of their
content.
Schemas perform a variety of functions. First, they lend organization
to an individual's experience in the sense that people order the elements
of their environment to reflect the structure of relevant schemas. Second,
schemas influence "what information will be encoded or retrieved from
memory" (Taylor and Crocker, 1981, p. 98). Third, the structure of a
schema constitutes a basis for "filling in" missing information and thus
HOW PEOPLE ORGANIZE THE POLITICAL WORLD 97
going beyond the information given (Rumelhart and Ortony, 1977).

Fourth, schemas provide a means for solving problems by supplying short
cuts or heuristics that simplify the problem-solving process. Finally, by
generating expectations against which reality is compared, schemas pro-
vide a basis for evaluating one's experiences (Taylor and Crocker, 1981).
It is also useful to have some understanding of a schema's structure.
In the most basic sense, the structure of a schema must define the
domain of relevant information and provide a means of organizing that
information in some consistent fashion (Fiske, 1981). One organizational
property that all schemas are expected to share is "a pyramidal structure,
hierarchically organized with more abstract or general information at the
top and categories of more specific information nested within the general
categories" (Taylor and Crocker, 1981, p. 92; also see: Cantor and
Mischel, 1979; Rumelhart and Ortony, 1977). Thus schemas process or
structure information at different levels of abstraction. In addition to a
hierarchical structure, the elements of a schema may also reflect other
organizational properties. For example, person schemas may be based on
categorical systems having to do with race, sex, or a variety of other
social categories, and balance may be an important organizing principle
for schemas heavily laden with affect, such as those involving interper-
sonal relations (Taylor and Crocker, 1981). Finally, it is important to
recognize that schemas are not necessarily isolated cognitive structures.
Rather, they may be linked with one another through a rich network of
hierarchical relationships in which individual schemas are "embedded"
in one another so that the higher-order, more abstract schemas are
characterized in terms of their more concrete, lower-order constituents
(Rumelhart and Ortony, 1977; Hastie, 1981).
To this point we have not yet considered affect-an element tradi-
tionally important to the study of political belief systems. In this regard,
it is critical to recognize that schemas are technically cognitive structures
for which neither affect nor value relevancy is a necessary precondition
for their activation (Fiske and Linville, 1980; Taylor and Crocker, 1981,
p. 125). As cognitive misers people often make inferences, judgments,
and predictions in what are essentially unvalenced situations. Yet, although
affect may not be essential to the functioning of schemas, neither is it
necessarily irrelevant. Some schemas-though certainly not all-represent
affectively laden structures of knowledge, and thus they are able to cue
affective, as well as cognitive, interpretations of a situation (Fiske, 1981).
Unfortunately, at this time the interplay between affect and cognition in
the functioning and structure of schemas is just beginning to come under
the scrutiny of social psychologists, and thus there is little direct empirical
evidence to guide our own efforts. Nonetheless, it is useful to note that
affect may well stimulate the development of schemas, heighten their
saliency, and influence their use (Fiske, 1982; Markus and Smith, 1981).
Schemas and the Study of Political Belief Systems
Our suggestion that the study of political belief systems be approached

from the perspective of schema theory is not a totally original one. Other
political scientists have also argued that the schema concept might
profitably be applied to the way people organize their political ideas
(Axelrod, 1973; Fiske and Kinder, 1981; Graber, 1982; Kinder, 1982;
Lodge and Wahlke, 1982; Sears and Citrin, 1982). Yet, missing from most
of these previous formulations is a comprehensive justification for
applying the concept specifically to mass belief systems and clear directives
as to how political schemas might be empirically studied. We address
each of these areas here.
Benefits of the Schema Concept
An application of schema theory does two things that Bennett (1977)

has suggested might be useful in alleviating the paradigmatic crisis that
currently characterizes the study of mass belief systems: it helps to bridge
the gaps among previous conceptualizations of the nature and structure
of mass belief systems, and it provides a theoretical basis for understand-
ing certain phenomena essentially beyond the confines of traditional
treatments. Specifically, the schema concept is by no means incompatible
with those previously used in conceptualizations of mass belief systems
(Conover and Feldman, 1980). In terms of the elements of a mass belief
system, an individual's more abstract or general schemas would be
analogous to what others have referred to as a "core belief system"
(Lane, 1973), "general ideological dimensions" (Jackson and Marcus,
1975), "general political orientations" (Thomas, 1978) and "ideological
principles" (Marcus et al., 1974). Similarly, while it is inappropriate to
think of a single-issue position as constituting a schema, it is reasonable
to conceptualize the perspective a person takes on related issues as an
emotionally laden, concrete schema toward a specific policy area.1 Thus
the general elements of a political belief system can be thought of in
terms of schemas that vary in their specificity and level of abstraction.
The structure of a belief system can also be described in schematic
terms. Belief system structure has traditionally been conceptualized in
terms of "constraint": the functional interdependence among idea ele-
ments (Converse, 1964). From the perspective of schema theory, a schema
in and of itself constitutes a "constrained" or organized set of informa-
tion. Thus, even when people's schemas are isolated from one another,
I Sears and Citrin (1982) suggest precisely this in their conceptualization of a tax-revolt
schema.
they can be said to have organized, albeit "atomized," ways of looking

at the political world.2
Finally, previous accounts of the development of belief systems are
also compatible with our model. Traditional sociological perspectives
e.g., Converse, 1964) emphasize the role of the environment in structuring
belief systems. In contrast, various "psychological" treatments (e.g.,
Lane, 1973) have focused on the manner in which internal mechanisms
structure beliefs. A schematic model integrates these two perspectives: a
schema develops as a consequence of interaction with the environment,
and subsequently as an existing internal structure it influences the way
new information is organized, thus shaping its own further development
(Neisser, 1976).
More important, not only does schema theory provide a conciliatory
view of the psychological and sociological conceptions of a political belief
system, it also broadens the range of phenomena likely to be dealt with
and understood within a belief system framework. In particular, the
schema concept is inextricably linked to models of the perceptual process
(Fiske and Linville, 1980, p. 546). This necessary link between structure
and process has been neglected-if not ignored-by those who study
political belief systems. Despite the fact that Converse (1964) cast his
seminal work in dynamic terms, political scientists have concentrated on
identifying the structure of belief systems without giving sufficient
attention to the full range of effects that structure might have on the
perceptual process. When political scientists have considered the percep-
tual effects of a belief system, they have focused almost exclusively on
evaluations: for example, how a voter's beliefs influence his or her
evaluation of a candidate. Yet, as is evident from our review of a
schema's functions, political schemas presumably have a range of effects
much wider than simply the structuring of evaluations. Thus one very
important consequence of adopting a schema theory approach is that it
provides a framework within which political scientists are likely to
consider a variety of linkages between the structure of beliefs and the
perceptual process.
A second consequence of applying schema theory to the study of
how people organize political beliefs is that it focuses attention on the
cognitive structure of political beliefs. Recall, although some schemas
may be tinged with affect, they are, nonetheless, cognitive structures of
knowledge. As such, they draw attention to those aspects of information
2This is not to suggest that all people have organized ways of looking at all facets of
the political world. Some people are "aschematic" (i.e., they lack organized beliefs) with
respect to certain areas of politics. But, saying that someone is aschematic is quite different
from arguing that they have an organized, but isolated, way of looking at that part of the
political world.
IOO Pamela Johnston Conover, Stanley Feldman
processing that are essentially cognitive or nonmotivational in nature.

People, in effect, make many judgments and inferences on strictly
cognitive grounds (Taylor and Crocker, 1981). In contrast, the traditional
approach to political belief systems concentrates heavily on the affective
structure of beliefs (Lodge and Wahlke, 1982), and the motivational
biases that may influence the impact of those beliefs on evaluations.
Thus a schematic view of the structure of political beliefs would result in
a greater consideration of the cognitive functions that belief systems
perform.
Finally, schema theory is also compatible with a more diverse, richer
view of the structure of political beliefs. From traditional perspectives, a
highly constrained belief system is expected to fit a relatively simple
model of bipolar, liberal-conservative structure. In contrast, schema
theory allows for several possibilities. First, liberal and conservative views
of politics may be structured very differently, rather than being bipolar
to one another (Conover and Feldman, 1981). Second, schema theory is
consistent with the idea of several distinct liberal (or conservative)
perspectives, each of which may focus on different aspects of the political
world. Third, in a more general sense, schema theory suggests that people
may employ a variety of organizing principles in structuring their beliefs.
Some may organize their schemas around party and race, while others
may order their political beliefs according to values such as "individual-
ism" and "equalitarianism" (Kinder, 1982). Finally, schema theory forces
us to broaden our notion of what constitutes a structured belief system;
people can have very organized, but atomized, ways of viewing the
political world.
Political Schemas
In applying schema theory to the study of political belief systems,

two important points must be kept in mind. First, while some aspects of
political knowledge may be relatively low in affect (e.g., ideas about how
Congress functions), those types of information relevant to political belief
systems are likely to be affectively loaded (Sears and Citrin, 1982, p. 76).
For example, schemas centered on concepts like "freedom" should be
heavily flavored by affect and therefore capable of triggering emotional
as well as cognitive reactions (Fiske, 1982). Moreover, such emotionally
laden schemas are expected to play an important role in explaining what
has come to be called "symbolic" politics (for an elaboration on the role
of schemas in symbolic politics, see Sears and Citrin, 1982).
Second, most of the social cognition literature deals with consensual
schemas-knowledge structures that people are expected to share. As a
consequence, relatively little attention has been paid to the individual
differences that may characterize the availability and use of schemas
(Fiske and Kinder, 1981). Yet there is likely to be a great deal of
HOW PEOPLE ORGANIZE THE POLITICAL WORLD IOI
variability between people in terms of the political schemas that they

possess and the manner in which such schemas are used (Fiske and
Kinder, 1981; Lodge and Wahlke, 1982). Thus any analysis of political
schemas must be careful to take into account the variation in the range
of political schemas that are available to particular individuals.
With this in mind, we can consider what form-in terms of content
and level of abstraction-political schemas are likely to take. First, with
respect to content, there are a variety of domains about which people
may have developed political schemas (e.g., the self, political parties, and
Congress). But we would argue that the domains most relevant in defining
a belief system are those that broadly concern the functioning of
government and the definition of public problems. In this regard, previous
research (e.g., Converse, 1964) indicates that domestic and foreign affairs
constitute two major domains of stimuli about which people have political
beliefs. Furthermore, with regard to the domestic domain, evidence
(Knoke, 1979) suggests that beliefs are structured according to whether
they involve economic matters, racial affairs, or social concerns. Thus a
schematic view of political belief systems should take into account
schemas that concern all four of these domains. Second, schemas differ
not only in their domains but also in their levels of abstraction. In
principle, it is possible that a person might have several schemas
concerning the same domain, but at different levels of abstraction.
Consequently, it is essential to examine political schemas at different
levels of abstraction.
Measuring Political Schemas
The empirical study of schemas has lagged behind theorizing in the

sense that there is no consensus on how schemas ought to be measured
(Fiske and Linville, 1980). Most social psychological studies have linked
the measurement of schemas to some aspect of information processing
(for a review of some of these methods see Ostrom, Pryor and Simpson,
1981). For example, in one measurement procedure people are given a
list of information to read and then later asked to recall what they have
seen. Presumably, the organization of the information in the subject's
recall will parallel its structure in their memory. Most recent studies of
political schemas (Fiske and Kinder, 1981; Lodge and Wahlke, 1982) have
tended to adopt this general approach (an exception is Sears and Citrin,
1982).
In contrast, there have been relatively few efforts to measure schemas
independent of their processing functions (Fiske and Linville, 1980;
Taylor, 1981; Taylor and Fiske, 1981). What studies there have been
(Cantor and Mischel, 1979; Markus, 1977; Markus and Smith, 1981)
suggest that one way of measuring political schemas independently of
information processing is for the researcher to define the possible range
of content (i.e., the elements) for each domain and at differing levels of
abstraction, and then allow the respondents to identify, through some
sort of rating task, the nature of the schemas that are most relevant to
their own particular way of looking at that portion of the political world.
We pursue this research strategy through the use of Q methodology.
Q methodology is a general approach to the study of attitudes,
beliefs, and preferences that is based on an examination of relationships
among people rather than among variables (for a review of Q method-
ology see Brown, 1980). Q analysis emphasizes the discovery of patterns
that characterize some subset of the respondents instead of examining
extent to which an hypothesized construct or relationship describes all
members of the sample. Theory or prior research guides the selection of
stimuli for the analysis, but the ways in which subjects can respond to
the stimuli are left largely unstructured. The central data-collection
instrument in Q methodology is the Q sort in which subjects are presented
with a large number of stimuli or statements from a particular domain.
They are then instructed to divide the statements into a number of
categories ranging from, for example, most strongly disagree through
neutral to most strongly agree. The next step is to compute a correlation
matrix among the Q sorts, or in other words, among people. To simplify
the data matrix and examine common patterns of sorting the stimuli, the
correlation matrix is factor analyzed and the initial factors rotated to a
final solution. Two strategies are available to interpret the results: the
factor loadings point to the individuals who best define the factors, and
the factor scores indicate which statements or stimuli are most character-
istic of each factor (Brown, 1980).
From our perspective, Q methodology provides an excellent means
of assessing political schemas. As researchers, we are able to specify the
possible range of political schemas by identifying the nature of stimuli
present in different domains and at varying levels of abstraction. The
individual is then allowed to reveal through the rating procedure the
manner in which he or she personally structures those stimuli. By asking
the respondents to rate the statements according to whether they agree
strongly, are neutral, or disagree strongly we have allowed them to reveal
both the affective and cognitive structure they lend to that domain of
stimuli: the critical dimensions of affective structure-valence and inten-
sity (Fiske, 1981)-are explicit in their ratings, while the essential elements
of cognitive organization are implicit in the overall pattern of the ratings.3
In addition, unlike more traditional ways of measuring beliefs Q meth-
3While we argue that Q methodology does reveal elements of both cognitive and
affective structure, it does not do so in a way that allows us to easily separate the two.
This is not a major concern for us, however, since our interest lies in assessing the overall
(both cognitive and affective) patterns of belief organization.
odology can identify the stimuli from a domain that are most salient to
an individual.
While it might be interesting to identify idiosyncratic forms of
structure, it is probably more useful to specify common patterns of
schematic organization since one function of patterns of political belief
is the communication of shared meaning in society. In this regard, the
factors represent a common way of structuring the stimuli from a specific
domain; in essence they define schemas shared among some portion of
the respondents. The nature of these shared schemas can be determined
by examining the factor scores and thereby identifying the statements or
stimuli most useful in defining the schema. Furthermore, given that the
factor loadings are the correlations between each individual Q sort and
the general pattern defined by the factor, they can be interpreted as
representing the extent to which a particular schema characterizes an
individual's responses to the stimuli making up that domain. Significant
positive loadings can be taken as an indication that the individual has
the shared schema defined by that factor; significant negative loadings
can be interpreted as meaning that the person has a schema that is the
mirror image of that defined by the factor. Finally, if a subject fails to
have a statistically significant loading on any of the factors derived for
that Q sort it would indicate that the person did not have a well-developed
schema for that area-at least not one that was shared by some other
subjects in the analysis. We use the term aschematic to refer to such an
individual.
Data
As an initial test of our schematic model of political belief systems,

we administered a set of six Q sorts to 59 students in introductory
political science courses. This sample of students had the following
demographic and political composition: 66/o male; 950o white; 38%o
upper middle class, 48%o middle class, 70o lower middle class, and 10%o
poor/working class (self-identification); 36%o attend religious services
weekly or more often, 43%o do so occasionally, and 210o never do so;
30%o liberal, 50/o moderate, and 20/o conservative; 350No Democratic,
280o independent, and 370o Republican; and finally, 520o do not read
the newspaper or watch the evening news regularly. Thus, while our
sample is relatively homogeneous in terms of social background (i.e.,
class and education), it is considerably more diverse in terms of political
ideology and interest. Clearly, respondents who differed more in their
social background would be necessary were we primarily interested in
examining the development of political schemas. However, given that our
focus is on studying the schematic structure of political beliefs, and given
that our sample does vary substantially on key political variables, this
group is adequate for an initial testing of our theory and methods. This
is especially true given that the representativeness of the respondents is

not nearly as crucial in Q methodology as it is in more traditional forms
of analysis. Much more important in Q methodology is the sample of
statements chosen to represent the possible stimuli within a particular
domain (Brown, 1980).
Six Q sorts were used to measure schematic structure. In each case,
they were based on a "forced distribution" that required respondents to
place a certain number of statements in each response category (e.g.,
"strongly agree"). Some may object that this creates a structure where
none exists. However, evidence suggests that (1) the specific shape of the
Q sort distribution makes little or no difference to the results, (2) even
forced choice distributions allow enormous freedom to sort statements in
an idiosyncratic manner (over 40 statements were used in each Q sort),
(3) there is evidence that forced distribution Q sorts have very high test-
retest reliabilities, and (4) the forced distribution encourages the respond-
ents to make distinctions among the stimuli that they are capable of but
may not recognize (Brown, 1980). In part, our analysis will help us to
resolve this issue since, if we have "manufactured" schemas where none
in fact exist, such structures will not necessarily be interrelated in a
meaningful fashion.
With this in mind, let us review the nature of the six Q sorts used
in the analysis. First, there are four Q sorts, each of which is designed
to represent one of the four major domains of content typically thought
to comprise political belief systems: economic, racial, social, and foreign
affairs. The statements in all four of these Q sorts were formulated at a
relatively "subordinate" level of abstraction; they deal with the nature
of social and individual problems in that domain and their implications
for public action. These four Q sorts, however, are not pitched at the
lowest possible level of abstraction; they do not concern specific policy
proposals or issues. Second, there is one Q sort with statements structured
to represent a middle level of abstraction in which basic ideological
principles and more general preferences for government action were
considered. This Q sort is not specific to a particular domain of content;
rather in its generality it encompasses the broader sphere of political
activity. Third, there is one Q sort designed to tap a "superordinate"
level of abstraction in which general beliefs about human nature and
social interaction are assessed. Taken together, these six Q sorts span the
various domains of content as well as the basic levels of abstraction a
typical political belief system might encompass. Finally, let us note that
the stimuli composing each Q sort were selected according to our own
notions of what belonged in the domain, as well as extensive examinations
of previous work.
The respondents were also asked about their positions on eleven

specific issues spanning all four domains of content.4 We interpret these
issue positions as representing the most specific, concrete elements of a
political belief system. Furthermore, while such issue positions do not
constitute schemas in and of themselves, they should be structured by an
individual's schemas.
The economic, social, and foreign-affairs Q sorts were given at one
session, while the racial, ideological principles, and basic human philos-
ophy Q sorts were administered at a session one week later. Contamination
across levels of abstraction was thus minimized, as was contamination
between the economic and racial Q sorts-two areas we thought might
be highly related. Finally, a questionnaire containing the issue questions
was self-administered by the students during a class period three weeks
apart from the Q sort sessions.
In analyzing the Q sorts, we employed a principal-components
solution in which the maximum off-diagonal correlation was used as an
initial communality estimate.5 In determining how many factors to rotate,
we used two criteria: the scree-test (Cattell, 1965) and the requirement
that there be at last two significant loadings on the factor (Brown, 1980).
Finally, the solutions were rotated obliquely, thus allowing correlated
factors to emerge where appropriate. In general, we found that oblique
rotations result in simpler patterns of loading then orthogonal ones.
Findings
Nature of the Political Schemas
Let us begin by considering the nature of the political schemas

revealed by our Q sort analysis. The factors that emerged for each Q
sort have been labeled according to our interpretations of the patterns
apparent in the factor scores (see Table 1). By examining those statements
that appear most positively and most negatively on a factor we were able
to infer the nature of the perspective or schema defined by each factor.
The Appendix provides a more detailed description of each schema as
well as a listing of statements with the most extreme factor scores.
First, for our purposes, perhaps the most important aspect of these
results is that for each combination of domain and level of abstraction
we uncovered at least two-and in some instances more-shared schemas
for structuring the same political information. In several domains, for
example, there are two distinctly different conservative perspectives, while
4The specific issue position items were drawn from National Election Studies of the
Center for Political Studies (CPS) and National Opinion Research Center (NORC) General
Social Surveys.
5 We used the Statistical Package for the Social Sciences (SPSS) factor program in our
analysis.
TABLE 1
The Nature of Political Schemas
/o oof N oof
Number Sample Sa
Loading Loading Loading
Schema on Factor Positive Negative Aschem
I.A.Basic human philosophy 15
Altruistic-positive (F1l) 29 47.5 1.7 r1
B. Hobbesian-Freudian (F12) 10 16.9 0 r1
C. Individualistic (F13) 11 18.6 0 r14
D. Altruistic-complex (F14) 6 8.5 1.7
II.A.Ideological principles 15.
Neoconservativism (F21) 30 49.2 1.7
B. Free-market conservativism (F22) 14 22.0
C. Democratic socialism (F23) 10 16.9 0
III. Economic beliefs 8.
A. Value of free enterprise (F31) 28 45.8 1
B. Society's responsibility for inequities (F32) 15 25
C. Value of self-reliance (F33) 17 27.1 1.
TABLE 1 continued
% of % of
Number Sample Sam
Loading Loading Loading 0
IV. Racial beliefs 5.1

Schema on Factor Positive Negative Aschem
A. Liberal integrationism (F41) 46 76.3
V. Social beliefs 6.8
B. Conservative integrationism (F42) 15 23.7 3.4
A. Religious-liberal sex roles (F51) 41 6

B. Liberal sex roles-nonreligious (F52) 27 40.7 5.1
VI.A. Internationalism-nonmilitaristic
Foreign-affairs beliefs 20.
(F61) 20 32.2
B. Nationalism-ethnocentric (F62) 20 33.9
C. Isolationism-nonmilitaristic (F63) 7 5.
io8 Pamela Johnston Conover, Stanley Feldman
in others there are viewpoints that mix liberal and conservative ideas.
Moreover, at least 100/ of the respondents (6 people) significantly load
on each factor, and in many instances fully a quarter of the sample share
the schema.6 Thus by no means are the schemas we have identified purely
idiosyncratic structures.
Second, an examination of the correlations among the factors within
a Q sort provides some idea of the distinctiveness of these schemas. A
high correlation among two factors indicates that the schemas defined by
the factors are related or share a good deal in common. Alternatively, a
low correlation between factors implies that the two schemas are relatively
independent; that is, they have very little meaning in common. As detailed
in Table 1, in most cases the correlations among the derived factors range
from being moderate to nonexistent. The one major exception to this
pattern occurs in the economic beliefs Q sort. There we find two related,
but distinct, conservative schemas (r = .44 for factors 1 and 3) as well
as one liberal schema. Thus in most Q sorts, the various schemas can be
described as being relatively independent or neutral with regard to one
another.
Third, there is only one really bipolar factor on which substantial
numbers of people load both positively and negatively. That factor-
F63-can be thought of as defining two schemas that are mirror images
of one another; one is an isolationism-nonmilitaristic schema while the
other one (defined by negative loadings on the factor) is an internation-
alism-militaristic schema. For the remaining factors, there are relatively
few negative loadings, suggesting that each schema defines a single,
distinct perspective (as opposed to two perspectives).7 This lack of
bipolarity is critical because it casts serious doubts on traditional concep-
tualizations of political belief systems, which tend to assume that people
structure their beliefs in terms of one or two bipolar structures-typically
labeled liberal-conservative dimensions (Asher, 1980). Furthermore, this
lack of bipolarity is quite consistent with the recent findings that people
do not view either political parties (Weisberg, 1980) or liberal-conservative
labels (Conover and Feldman, 1981) from a bipolar perspective.
Fourth, it is interesting to note that for most of the Q sorts there
are a number of individuals who do not load on any factor (i.e.,
aschematics). They are lowest for the Q sorts dealing with what one
would expect to be very familiar information: social and racial beliefs.
Similarly, the number of aschematics is highest in an area where
6 Several people in each Q sort had significant loadings on two of the extracted factors.
This indicates that their schema for that domain combines two different ways of structuring
the relevant political information and beliefs.
7 This lack of bipolarity is also evident in the correlations among factors: there is only
one negative correlation, which is not sizeable.
HOW PEOPLE ORGANIZE THE POLITICAL WORLD 1O9
Americans are notorious for their lack of information and well-developed

beliefs: foreign affairs. Finally, it is important to recognize that the
tendency to be aschematic does not appear to be a general one; only 12
percent of the subjects are aschematic in more than one area.8
In summary, our analysis indicates that people who are by no means
political elites do have identifiable, shared ways of organizing political
information at abstract as well as relatively concrete levels. Furthermore,
within any given domain and level of abstraction, there are several
distinct, relatively independent schemas or ways of structuring the same
information, that often cannot be labeled in simple liberal-conservative
terms. These facts undermine those theories that depict the structure of
political beliefs strictly in terms of bipolar, liberal-conservative dimen-
sions. People simply do not view the political world from opposite sides
of the same dimension, nor do they necessarily see it in liberal-conserv-
ative terms. Instead, they bring distinct and varied perspectives to bear
on the political information that they receive. Finally, our results should
not be interpreted as identifying the full range of political schemas.
Future studies involving different kinds of people are expected to uncover
schemas not employed by our subjects.9
Structure of the Political Schemas
To this point, we have demonstrated that individuals do have different

schemas that they use to structure certain types of information. This
does not indicate, however, that people have integrated belief systems;
rather, all that it shows is that individuals have atomized perspectives.
Yet for many people this may be more than enough for them to structure
the limited amounts of political information that they encounter in their
daily lives. Much of the mass public simply may not need a broad-based
ideology in order to organize their political worlds (Kinder, 1982). In
contrast, others may require more integrated structures for processing
political information. Thus it is useful to illustrate that some people do
relate their schemas to one another.
8Some caution must be used in interpreting the aschematic percentage. Technically it

means only that people do not have a schema that is shared by other respondents in our
analysis. They still might have a very idiosyncratic schema for dealing with the information.
Test-retest data would provide one way of distinguishing the truly aschematic respondents
from those with simply idiosyncratic patterns of organization. Presumably, those with
idiosyncratic structures would correlate with themselves over time, while aschematics would
not.
9Similarly, a more diverse sample might also be useful in exploring the determinants
of various schemas. Given the relative homogeneity of our sample, such an investigation
would not have been very fruitful in this case. However, even a preliminary analysis using
this sample does indicate that people holding different schemas do differ significantly in
the groups they identify with and the values they hold.
Iio Pamela Johnston Conover, Stanley Feldman
In order to do so, however, it is necessary to shift from considering

the individual Q sorts to a focus on their interrelationships: a move that
leads us to depart from traditional Q sort analysis. Specifically, by
correlating an individual's factor loadings across factors we gain some
indication of the shared patterns of structure characterizing relations
between schemas.10 A significant positive correlation indicates that a
person having the first schema also tends to have the second one; a
nonsignificant correlation means that the schemas are relatively independ-
ent; and a significant negative correlation can be interpreted as meaning
that a person who holds the first schema tends not to have the second
one.11 Given our interest in identifying patterns of schematic structure,
we will focus primarily on the positive correlations since they indicate
which schemas tend to go with one another.
Presented in Table 2 are the correlations between the factor loadings
of the respondents on each of the seventeen factors or schemas we have
identified. Let us first note that the pattern of correlations between
domains (as defined by the Q sorts) varies depending on the particular
schema a person holds. For example, looking at the pattern of correlations
between the economic and the racial beliefs schemas, we find that, for
individuals having the "value of free enterprise" schema (F31), economic
and racial beliefs are relatively independent of one another. In contrast,
for people who organize their economic beliefs differently, the economic
and racial domains are clearly linked to one another; subjects with a
"society's responsibility for inequities" schema (F32) tend to have a
"liberal integrationism" schema (F41), while those with a "value of self-
reliance" schema (F33) tend to view racial matters from the perspective
of a "conservative integrationists" schema (F42). To take a second
example, for those with a "free-market conservatism" schema (F22)
ideological principles are essentially independent of the schemas used to
organize both racial matters and foreign affairs, while there are clear
relationships between the two domains for subjects who structure their
ideological principles differently.
This finding that the relationships across domains often vary depend-
ing on an individual's particular schemas has important implications.
Methodologically, it suggests that it may be quite misleading to simply
look at the relationships among attitudes in two domains. Because people
relate domains together in different ways, substantial relationships may
10 Using correlations between factor loadings as a measure of schematic structure makes

sense only if (as we have done) the factors being correlated are from different Q sorts.
Correlations between factor loadings for factors from the same Q sort would produce
spurious results.
11 The interpretation of a significant negative correlation differs where F63-the one
bipolar factor-is concerned. There a negative correlation means that people who have the
"mirror image" of F63 (those who load negatively on F63) tend to have the second schema.
TABLE 2
The Structure of Political Schemas
Ideologic
Basic Human Philosophy Principles Eco
Schema F11 F12 F13 F14 F21 F22 F23 F31 F32 F3
1. Basic human philosophy
A. Altruistic-positive (Fll)
B. Hobbesian-Freudian (F12)
C. Individualistic (F13)
D. Altruistic-complex (F14)
II. Ideological principles

A. Neoconservativism (F21) .30** -.21* .30** -.16
B. Free-market conservativism (F22) -.14 .23** .07 -.37**
C. Democratic socialism (F23) .19* -.06 -.49** .30**
III. Economic beliefs
A. Value of free enterprise (F31) .04 .25** -.14 -.31** .12 .51** -.34**
B. Society's responsibility for inequities (F32) .25** -.17* -.21** .50** -.13 -.33** .64**
C. Value of self-reliance (F33) -.18* -.06 .53" - .27** .32*' .04 -.55**
TABLE 2 continued
Ideologica
Basic Human Philosophy Principles Econom
Schema Fll F12 F13 F14 F21 F22 F23 F31 F32 F33
IV. Racial beliefs
A. Liberal integrationism (F41) .27** -.11 -.31** .04 -.28** .07 .38** -.08 .39** -.3
B. Conservative integrationism (F42) -.19* .04 .43* -.13 .28** .06 -.52** .11 -.41**
V. Social beliefs
A. Religious-liberal sex roles (F51) .30** -.32* .12 .40** .32** -.40** -.10 -.21** .34
B. Liberal sex roles-nonreligious (F52) -.03 .05 -.22** -.26** -.17* .06 .38** -.08 .11
VI. Foreign-affairs beliefs

A. Internationalism-nonmilitaristic (F61) .27** -.17* -.35** .25* -.13 -.12 .60** -.14 .48** -.5
B. Nationalism-ethnocentric (F62) -.16 .30** .20* -.14 .13 .06 -.48** .04 -.31 ** .54
C. Isolationism-nonmilitaristic (F63) -.06 -.14 -.03 .08 -.21* -.16 .21* -.38*
NOTE: Entries are Pearson product-moment correlations.
* = (p < .1). **(p < .05).
HOW PEOPLE ORGANIZE THE POLITICAL WORLD II3
be hidden or washed out when such differences are ignored. Theoretically,

this finding indicates that people have very different ways of looking at
politics. Some people see two domains of politics from the perspective of
relatively independent schemas; others with different schemas will view
the same domains as being quite related.
Moreover, not only do people relate the various domains of political
information in different ways, but they also do so in a substantively
meaningful fashion and a structurally hierarchial pattern. If we focus on
the relationships of the superordinate schemas with those identified at
both the middle and the subordinate levels, then we find evidence of a
hierarchial structure in which the specific schemas are embedded in the
more general, higher-level ones. For example, one generally conservative
perspective is illustrated by people who view human nature from the
perspective of a "Hobbesian-Freudian" (F12) schema. They tend to adopt
a "free-market conservativism" (F22) view when dealing with general
ideological information, and at the level of more specific beliefs they
organize economic information in terms of the "value of free enterprise"
(F31), and foreign affairs information in terms of a "nationalism-
ethnocentric" (F62) schema. In contrast, a second, distinctly different
conservative perspective is found among those individuals having an
"individualistic" schema (F13) at the superordinate level. Finally, the
embedding pattern that characterizes the relationships among schemas
becomes even more apparent when we move down a level and look at
how the "ideological principles" schemas structure more specific infor-
mation.
It is also of some use to consider the negative correlations in Table
2 since they indicate which schemas are not likely to be found together
in the same belief structure. When such correlations are examined the
patterns of hierarchical structure become clearer and the substantive
nature of the perspectives defined by the schemas more apparent. For
example, the basically liberal perspective of those with a "democratic
socialism" (F23) schema is made even more distinct by the recognition
that such people definitely tend not to structure their economic beliefs in
terms of the values of either "free enterprise" (F31) or "self-reliance"
(F33); nor do they order their racial beliefs according to a "conservative
integrationists" (F42) perspective, or their foreign affairs beliefs in terms
of a "nationalism-ethnocentric" (F62) schema. Similarly, the essentially
conservative perspective of those viewing human nature in "individualis-
tic" (F13) terms becomes quite clear when the negative correlations in
each of the other domains are taken into account.
In summary, not only do individuals have a wide variety of schemas,
but many people also link their schemas together in what appears to be
a meaningful hierarchial fashion. Overall, there is evidence of two
basically conservative-but decidedly different-perspectives, one essen-
I14 Pamela Johnston Conover, Stanley Feldman
tially liberal orientation, and several viewpoints that mix liberal and
conservative ideas. Furthermore, people relate schemas to one another in
very different ways, depending on the nature of their schemas. For some
people, two domains may be very related to one another, while for others
they are quite independent.
Schemas and Specific Issue Positions

To this point, our analysis has neglected two important elements. We
have given no consideration to specific issue positions-the elements of
a belief system that are often the basic unit of analysis in traditional
examinations of political belief systems. Nor have we directed any
attention to the functions that political schemas perform-this despite
our own argument that process and structure ought to be considered
together. In this section, we attempt to rectify both of these omissions
by considering how various schemas are related to specific issue positions.
While this analysis does not represent a true illustration of the dynamic
functioning of political schemas, it is certainly more than what some
might call a simple tautological demonstration. By showing that the
structure of schemas is in fact related to the evaluation of more specific
issues in the same domain, we lend considerable support to the idea that
our Q sort analysis has actually uncovered meaningful patterns of
schematic structure.
The subjects' factor loadings on the 17 schemas were correlated with
their issue positions on four issues, one from each of the major domains
of political content: should the government improve living standards,
school integration, marijuana use, and defense spending (see Table 3).12
In each case, the issue scales range from a low of 1 for what would
typically be considered the liberal response to a high of 7 for the most
extreme conservative response: people should take care of themselves, the
government should stay out of school integration, penalties should be
higher for marijuana use, and defense spending should be increased. In
this analysis, significant positive correlations can be interpreted as mean-
ing that individuals who hold a particular schema tend to adopt the
conservative position on specific issues, while a significant negative
correlation indicates that people with the schema tend to take a liberal
stand on specific issues.
People's political schemas are definitely related to the positions that
they take on specific issues as we see in Table 3. To begin with,
conservative issue positions tend to be associated with schemas that
represent a conservative perspective on some domain of political infor-
12 The four issues were chosen because they tend to represent the middle range of
correlation; they are not the issues most correlated with the schemas, but neither are they
the ones least correlated.
TABLE 3
The Structure of Political Schemas and Specific Issue Po
Government Improv
Schema Standard of Living Integration
I. Basic human philosophy
A. Altruistic-positive (Fll) -.17 -
B. Hobbesian-Freudian (F12) .0
C. Individualistic (F13) .26**
D. Altruistic-complex (F14) -.43** -
II. Ideological principles
A. Neoconservativism (F21) .00 .1
B. Free-market conservativism (F22)
C. Democratic socialism (F23) -.55* 3
III. Economic beliefs
A. Value of free enterprise (F31) .33
B. Society's responsibility for inequities (F32) -.65
C. Value of self-reliance (F33) .53** .
TABLE 3 continued
Government Improv
Schema Standard of Living Integration
IV. Racial beliefs
A. Liberal integrationism (F41) -.35**
B. Conservative integrationism (F42) .4
V. Social beliefs
A. Religious-liberal sex roles (F51) -.23
B. Liberal sex roles-nonreligious (F52) -.1
A. Internationalism-nonmilitaristic (F61) -.53*
B. Nationalism-ethnocentric (F62) *34* .4
C. Isolationism-nonmilitaristic (F63) -.21* -
NOTE: Entries are Pearson product-moment correlations.
* = (p < .1). ** = (p - .05).
HOW PEOPLE ORGANIZE THE POLITICAL WORLD II7
mation, while liberal stands on the issues are related to so-called liberal
schemas. Next, the pattern of significant relationships for a particular
schema tends to be consistent across issues (i.e., liberal schemas tend to
be associated with liberal issue positions). The one major exception to
this pattern is the religious/liberal sex-roles schema; people having that
schema tend to have a liberal position on the standard-of-living issue and
a conservative position on marijuana use. Yet such a pattern makes sense
when it is recognized that this schema combines a conservative religious
perspective with a liberal view of some, but not all, social matters.
Finally, those schemas most relevant to the domain of an issue tend
to have some of the strongest correlations with the issue positions. Thus,
for example, the three economic belief schemas are all strongly related
to stands on the standard-of-living issue, and the racial beliefs schemas
are the strongest correlates of positions on school integration. This is not
to say, however, that there are not strong correlations between certain
schemas and issues basically outside their domain. Indeed, there are
substantial correlations between the foreign affairs schemas and all three
domestic issues. On one hand, such correlations may indicate real
substantive links between domains traditionally thought to be independ-
ent. It is certainly plausible that basic beliefs about our own culture vis-
a-vis others (i.e., feelings of nationalism and ethnocentrism) may well be
related to issues of race and economics within our society. On the other
hand, these correlations may be spurious ones created by the relationship
of the foreign affairs schemas to other very basic schemas. To test this
possibility, however, we must shift our attention away from the structural
linkages between political schemas and specific issue positions to a
multivariate analysis and a more traditional focus on our ability to predict
specific issue positions.
We ran stepwise regression analyses in which we regressed the
subjects' stands on the 11 issues on their factor loadings for the 17
schemas (see Table 4). As can be seen, we do very well in predicting
specific issue positions for most of the economic issues and the defense
spending issue; the multiple R's are quite large for those issues, particu-
larly when one takes into account the relatively low reliabilities of the
issue questions.13 The schemas do slightly less well in predicting positions
on the two racial issues, and the least well of all on the four social issues.
But this poor performance on the social issues is not altogether unex-
pected; because the subjects are relatively similar for many of the
13 Analysis by Erikson (1979), for one, indicates that the issue-position scales may
contain a substantial degree of random measurement error. Therefore if an issue scale had
a reliability of .7, for example, an estimated R2 from a regression on that scale of .4 (40o1
explained variance) would actually represent .4/.7, or 57% of the systematic variation in
the issue scale explained.
TABLE 4
Multiple Regression Analysis: Specific Issue Positions on Polit
Govt. Standard Health Income Sch. Preferential Mari

Schema Activity of Living Insur. Differences Integr. Treatment Use Accuse
A. Altruistic-positive (F1l) .30

1. Basic human philosophy
B. Hobbesian-Freudian (F12) -.207

C. Individualistic (F13)
D. Altruistic-complex (F14) -.225
11. Ideological principles
A. Neoconservativism (F21)
B. Free-market conservativism (F2) .569 .33
C. Democratic socialism (F23)
Ill. Economic beliefs
inequities (F32) -.331 -

A. Value of free enterprise (F31)
B. Society's responsibility for
C. Value of self-reliance (F33) .226
TABLE 4 continued
Govt. Standard Health Income Sch. Preferential Mari

Schema Activity of Living Insur. Differences Integr. Treatment Use Accuse
IV. Racial beliefs
A. Liberal integrationism (F41) -.324
B. Conservative integrationism (F42) .476 .523 .401
A. Religious-liberal
V. Social beliefs
B. Liberal sex roles (F5(

sex roles-nonreligious
A. Internationalism-nonmilitaristic (F61) -.219
B. Nationalism-ethnocentric (F62) .41
C. Isolationism-nonmilitaristic
Multiple R .68 .78 .66 .59 .52 .64 .4
Multiple R2 .47 .61 .43 .35 .27 .40 .22
NOTE: Entries are standardized regression coefficients significant at the .05 level.
dimensions of social beliefs, the two social beliefs schemas overlap

considerably, thus reducing our ability to discriminate among people for
issues in that domain. Nonetheless, despite the relatively low level of
explanation for the social issues, overall the schemas do very well in
predicting specific issue positions.
Furthermore, we again find that "liberal" schemas tend to contribute
to liberal issue positions and vice versa, with only a few apparent
exceptions. First, those with "free-market conservativism" schemas (F22)
tend to oppose increases in defense spending. However, given the emphasis
that we would expect such individuals to place on reduced governmental
spending and a balanced budget, such a finding is not especially incon-
sistent with how we would expect schemas to structure issue positions. A
second, more difficult anomaly to reconcile is the finding that people
with a "Hobbesian-Freudian" view tend to support a federal health
insurance program. This link inexplicably goes against the basically
conservative perspective characterizing these subjects on other issues and
schemas.
Looking at the domain of the schemas, we generally find further
support for our earlier analysis. Again, the schemas most useful for
predicting issue positions tend to be those most germane to the domain
of the specific issue. For example, the racial schemas are the best
predictors of positions on the two racial questions. Similarly, the eco-
nomic isues are best predicted from the general ideological and economic
schemas, with the racial schemas playing a role on several of the issues.
However, the influence of the racial schemas on economic issue positions
is quite consistent with recent research indicating the centrality of race to
American belief systems (see Kinder, 1982). The one interesting exception
seems to be social issues where we find that the foreign affairs schemas
are relatively strong predictors. In effect, they seem to be picking up
differences in social outlooks not identified by the social beliefs schemas.
Moreover, as suggested earlier, this finding raises the possibility that there
are important substantive linkages between basic beliefs on foreign affairs
and those concerning other domains. At a fundamental level, then, our
beliefs about other cultures may very well be tied to our beliefs about
other races and our sense of morality.
In summary, the schemas that we have identified are related to
specific issue positions-the most concrete elements of a political belief
system-in meaningful ways. While true experiments demonstrating the
information-processing effects of schemas might be more persuasive,
these findings can nonetheless be interpreted as evidence that schemas do
structure the evaluation of specific policies. In addition, from a more
traditional perspective, we find that we can use information about whether
or not people have particular schemas to successfully predict their
positions on specific issues.
Conclusions
In this paper we have both developed a schematic model to explain

the ways in which people organize the political world and tested a general
strategy for the empirical study of political schemas. We have argued
that schema theory provides a useful basis for the study of mass belief
systems, in particular, because it allows for a more complete specification
of the diverse structures assumed by political beliefs. Our empirical
analysis using Q methodology found consistent support for the utility of
a schematic model of political belief systems. In each of the six domains
analyzed, we found multiple, distinct schemas representing different ways
of organizing information in that substantive area. Some of these were
liberal schemas and others were conservative ones, but the liberal per-
spectives in a domain were not simply reflections of the conservative
viewpoints. Moreover, for most of our subjects the schemas we identified
were related to each other in ways indicative of substantial belief system
organization. While we would hesitate to label such organization as an
ideology, we readily interpret it as evidence that fairly average people
may have relatively complex, interrelated ways of structuring their political
world. These findings suggest that people organize their political worlds
in richer and more diverse ways than implied by the traditional approaches
to mass belief systems.
In conclusion, we believe that a schematic model of the organization
of political information holds a great deal of promise. People may not
necessarily structure their political worlds according to abstract ideological
principles, but they do organize their beliefs. Furthermore, they do so in
many ways. It is no longer sufficient to look for elements of a simple
liberal-conservative structure; rather, as researchers, we must recognize
that there are a number of distinct, often unrelated, perspectives on
politics. Specifying the nature of those perspectives and exploring their
impact on political perception represents an important, and challenging,
research agenda for the future.
Manuscript submitted 25 October 1982

Final manuscript received 27 June 1983
APPENDIX
Listed are the 17 schemas and, for illustrative purposes, the two statements with the
most extreme positive-factor scores and the two statements with the most extreme negative
scores. It is critical to remember that the naming of the factors was based on the full array
of positive and negative rankings of statements; in some instances the meaning of the
factors may not be at all clear from the limited number of statements presented. A complete
record on the results of the six factor analyses can be obtained by request from the authors.
1. BASIC HUMAN PHILOSOPHY
A. Altruistic-positive: People are seen as being basically good, trusting, and willing to help
each other; altruism is considered important.
+ 1. You can't put a price tag on human life.
+2. If you act in good faith toward people, most all of them will reciprocate with fairness
toward you.
- 1. All in all, it is better to be important and dishonest than to be humble and honest.
-2. An individual's responsibility for the welfare of others extends no further than the
boundaries of his or her immediate circle of friends.
B. Hobbesian-Freudian: A view of people as self-interested, untrustworthy, and difficult
to understand.
+ 1. It's a rare person who will go against the crowd.
+ 2. People are too complex to ever be fully understood.
- 1. The average person has an accurate understanding of the reasons for their behavior.
- 2. People usually tell the truth, even when they know they would be better off lying.
C. Individualistic: Success and failure in life is a matter of personal initiative; people
deserve what they get.
+ 1. If people try hard enough they can usually reach their goals.
-2. Our success in life is pretty much determined by forces outside our own control.
D. Altruistic-complex: People are basically self-interested and difficult to understand, but
they still have an obligation to the community as a whole.
+ 2. Honesty is the best policy in all cases.
- 2. The average person has an accurate understanding of the reasons for their behavior.
II. IDEOLOGICAL PRINCIPLES
A. Neoconservativism: Big government is seen as a necessary evil needed to maintain law
and order; existing customs and institutions are important.
+ 1. The maintenance of law and order is essential to the sound development of society.
+ 2. Society can be improved by ideas.
- 1. There should be no interference with business and trade.
-2. A better society can only be realized through a radical change of the present social
structure.
B. Free-market conservativism. A basic antigovernment view that stresses the need for
freedom in order to pursue business interests.
+ 1. A great deal of government interference can only lead to bureaucracy and economic
stagnation.
+ 2. If freedom of enterprise is restricted other freedoms will disappear.
- 1. A great deal of government interference leads to planning and therefore a more
efficient economy.
- 2. Efficient, large-scale production requires government intervention.
C. Democratic socialism. A view in which change is perceived as a necessary and positive
element in society. The existence of social classes is viewed as a detriment to society.
+ 1. Society can be improved by ideas.
+ 2. The maintenance of law and order is essential to the sound development of society.
- 1. The existence of social classes is necessary for the welfare of all.
-2. In present-day society, social classes no longer form an important social conflict.
III. ECONOMIC BELIEFS
A. Value of free enterprise. A focus on the positive value of competition, big business, and
profit-making.
+ 1. Private ownership of property is as important to a good society as freedom.
HOW PEOPLE ORGANIZE THE POLITICAL WORLD I23
+ 2. Competition leads to better performance and a desire for excellence.

- 1. People would still work as hard at their jobs even if everyone earned the same amount.
-2. Private ownership of property has often done mankind more harm than good.
B. Society's responsibility for inequities: Antibusiness viewpoint that sees society as sharing
in the responsibility for social inequities.
+ 1. Too often in society, success is defined just in terms of how much money you make.
+2. Society needs to work harder to ensure real equality of opportunity.
- 1. Currently opportunities for advancement are about as equal as they need to be.
-2. Business and industry are generally fair and honest with the public.
C. Value of self-reliance: A conservative view that focuses on the individual causes of
poverty.
+ 1. Some people who don't get ahead in life tend to blame the system, when they really
have only themselves to fault.
+ 2. Competition leads to better performance and a desire for excellence.
- 1. The poor are poor because the wealthy and powerful keep them poor.
-2. Competition, whether in school, work, or business is often wasteful and destructive.
IV. RACIAL BELIEFS
A. Liberal integrationism. A view that favors integration at both a public (social) level as
well as at a personal (individual) level.
+ 1. There is nothing wrong with blacks and whites being close personal friends
+ 2. It is certainly proper for blacks and whites to be acquaintances.
- 1. A person should not invite a member of another race to dinner at his or her home.
-2. There should be no blacks serving on the city council.
B. Conservative integrationists. A view that supports integration at the public level, but
opposes it at a close personal level. Also, there is some tendency not to perceive the
existence of discrimination.
+ 1. People should be willing to take orders from a black police officer.
+ 2. It is certainly proper for blacks and whites to be acquaintances.
- 1. It is quite all right for blacks and whites to date each other.
-2. When two qualified people, one black and the other white, are considered for the
same job, the black won't get the job no matter how hard he or she tries.
V. SOCIAL BELIEFS
A. Religious-liberal sex roles. An emphasis on the importance of religious faith, with some
tendency towards liberal sexual attitudes.
+ 1. There is nothing wrong with a married woman working even if she has a husband
capable of supporting her.
+ 2. Religious commitment gives life a purpose it would not otherwise have.
- 1. The Bible is not the actual word of God; it is simply an ancient book of fables,
legends, and history recorded by humans.
- 2. There is no survival of any kind after death.
B. Liberal sex roles-nonreligious. Definitely nonreligious with an emphasis on liberal sexual
attitudes.
+ 1. There is no reason why a man should lose respect for a woman if they have sexual
relations before marriage.
+2. If you lead a good and decent life it is not necessary to go to church.
- 1. Every explanation of man and the world is incomplete unless it takes account of
God's will.
-2. The story of creation as recorded in Genesis is literally true.
VI. FOREIGN AFFAIRS BELIEFS
A. Internationalism-nonmilitaristic. A stress on world-as opposed to U.S.-interests and
the avoidance of war whenever possible.
+ 1. All human beings are of equal importance.
+2. International disputes should be settled without war.
- 1. We should be willing to fight for our country whether it is in the right or wrong.
-2. The best way to ensure peace is through military strength.
B. Nationalism-ethnocentric. A basically nationalistic view with some militaristic overtones.
+ 1. It is only natural and right for people to feel that their country is better than any
other.
+ 2. All human beings are of equal importance.
- 1. Our country is probably no better than many others.
-2. The United States should limit itself to defensive weapons only.
C. Isolationism-nonmilitaristic. A heavy emphasis on staying uninvolved in world affairs,
with nonmilitaristic overtones.
+ 1. George Washington's advice to stay out of agreements with foreign powers is just as
wise now as it was when he was alive.
+ 2. We shouldn't risk our happiness and well-being by getting involved with other countries.
- 1. The United States should go out of its way to lend a helping hand to all countries.
- 2. The biblical command against killing does not apply to warfare.
REFERENCES
Asher, Herbert B. 1980. Presidential elections and American politics. 2nd ed. Homewood,
Ill.: Dorsey.
Aaxelrod, Robert. 1973. Schema theory: An information processing model of perception
and cognition. American Political Science Review 67 (December 1973): 1248-66.
Bennett, W. Lance. 1977. The growth of knowledge in mass belief studies: An epistemolog-
ical critique. American Journal of Political Science 21 (August 1977): 465-500.
Brown, Steven R. 1980. Political subjectivity: Applications of Q methodology in political
science. New Haven, Conn.: Yale University Press.
Cantor, Nancy, and Walter Mischel. 1979. Prototypes in person perception. In Advances in
experimental social psychology, vol. 12. edited by Leonard Berkowitz, pp. 3-52. New
York: Academic.
Cattell, R. B. 1965. Factor analysis: An introduction to essentials. (I) the purpose and
underlying models, (II) the role of factor analysis in research. Biometrics 21 (March-
June 1965): 190-215, 405-35.
Conover, Pamela Johnston, and Stanley Feldman. 1980. Belief system organization in the
American electorate: An alternate approach. In The electorate reconsidered, edited by
John C. Pierce and John L. Sullivan, pp. 49-67. Beverly Hills, Calif.: Sage.
. 1981. The origins and meaning of liberal/conservative self-identifications. American
Journal of Political Science 25 (November 1981): 617-45.
Converse, Philip E. 1964. The nature of belief systems in mass publics. In Ideology and
Discontent, edited by David Apter, pp. 202-61. New York: John Wiley.
Erikson, Robert S. 1979. The SRC panel data and mass political attitudes. British Journal
of Political Science 9 (January 1979): 89-114.
Fiske, Susan T. 1981. Social cognition and affect. In Cognition, social behavior and the
environment, edited by John H. Harvey, pp. 227-64. Hillsdale, N.J.: Lawrence
Erlbaum.
. 1982. Schema-triggered affect: Applications to social perception. In Affect and
cognition, edited by Margaret Syndor Clark and Susan T. Fiske, pp. 55-78. Hillsdale,
N.J.: Lawrence Erlbaum.
Fiske, Susan T., and Donald R. Kinder. 1981. Involvement, expertise, and schema use:
Evidence from political cognition. In Personality, Cognition, and Social Interaction,
edited by Nancy Cantor and John F. Kihlstrom, pp. 171-90. Hillsdale, N.J.: Lawrence
Erlbaum.
Fiske, Susan T., and Patricia W. Linville. 1980. What does the schema concept buy us?
Personality and Social Psychology Bulletin, 6 (December 1980): 543-57.
Graber, Doris A. 1982. Have I heard this before and is it worth knowing?: Variations in
political information processing. Paper prepared for delivery at the 1982 Annual
Meeting of the American Political Science Association, Denver, Colorado, September
1-5.
Hastie, Reid. 1981. Schematic principles in human memory. In Social cognition: The
Ontario symposium, vol. 1., edited by E. Tory Higgins, C. Peter Herman, and Mark
P. Zanna, pp. 39-88. Hillsdale, N.J.: Lawrence Erlbaum.
Jackson, Thomas H., and George E. Marcus. 1975. Political Competence and Ideological
Constraint. Social Science Research 4 (June 1975): 93-111.
Kinder, Donald R. 1982. Enough already about ideology: The many bases of American
public opinion. Paper prepared for delivery at the 1982 Annual Meeting of the
American Political Science Association, Denver, Colorado, September 1-4.
Knoke, David. 1979. Stratification and the dimensions of American political orientations.
American Journal of Political Science 23 (November 1979): 772-91.
Lane, Robert E. 1973. Patterns of political belief. In Handbook of Political Psychology,
edited by Jeanne Knutson, pp. 83-116. San Francisco: Jossey Bass.
Lodge, Milton, and John C. Wahlke. 1982. Politicos, apoliticos, and the processing of
political information. International Political Science Review 3 (January 1982): 131-50.
Marcus, George, David Tabb, and John L. Sullivan. 1974. The application of individual
differences scaling to the measurement of political ideologues. American Journal of
Political Science 18 (May 1974): 405-20.
Markus, Hazel. 1977. Self-schemata and processing information about the self. Journal of
Personality and Social Psychology 35 (February 1977): 63-78.
Markus, Hazel, and Jeanne Smith. 1981. The influence of self-schemas on the perception
of others. In Personality, cognition, and social interaction, edited by Nancy Cantor
and John H. Kihlstrom, pp. 233-62. Hillsdale, N.J.: Lawrence Erlbaum.
Neisser, Ulric. 1976. Cognition and reality: Principles and implications of cognitive
psychology. San Francisco: W. H. Freeman.
Nie, Norman, Sidney Verba, and John R. Petrocik. 1976. The changing American voter.
Cambridge, Mass.: Harvard University Press.
Ostrom, Thomas M., John B. Pryor, and David D. Simpson. 1981. The organization of
social information. In Social cognition: The Ontario symposium, vol. 1, edited by E.
Tory Higgins, C. Peter Herman, and Mark P. Zanna, pp. 3-38. Hillsdale, N.J.:
Lawrence Erlbaum.
Rumelhart, David E., and Andrew Ortony. 1977. The representation of knowledge in
memory. In Schooling and the acquisition of knowledge, edited by R. C. Anderson,
R. J. Spiro, and W. E. Montague, pp. 99-135. Hillsdale, N.J.: Lawrence Erlbaum.
Sears, David O., and Jack Citrin. 1982. Tax revolt: Something for nothing in California.
Cambridge, Mass.: Harvard University Press.
Taylor, Shelley E. 1981. The interface of cognitive and social psychology. In Cognition,
social behavior and the environment, edited by John H. Harvey, pp. 189-211. Hillsdale,
N.J.: Lawrence Erlbaum.
Taylor, Shelley E., and Jennifer Crocker. 1981. Schematic bases of social information
processing. In Social cognition: The Ontario symposium, vol. 1, edited by E. Tory
Higgins, C. Peter Herman, and Mark P. Zanna, pp. 89-134. Hillsdale, N.J.: Lawrence
Erlbaum.
Taylor, Shelley E. and Susan T. Fiske. 1981. Getting inside the head: Methodologies for
process analysis in attribution and social cognition. In New Directions in Attribution
Research, vol. 3, pp. 459-524, edited by J. H. Harvey, W. Ickes and R. F. Kidd.
Hillsdale, N.J.: Lawrence Erlbaum.
Thomas, Dan B. 1978. Political belief systems and ideo-affective resonance: The structuring
principle revisited. Experimental Study of Politics 6 (1978): 34-89.
Weisberg, Herbert. 1980. A multidimensional conceptualization of party identification.
Political Behavior 2 (1980): 33-60.

Conover PeopleOrganizePolitical 1984

Uploaded by

Copyright:

Available Formats

Conover PeopleOrganizePolitical 1984

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Conover PeopleOrganizePolitical 1984

Uploaded by

Copyright:

Available Formats

How People Organize the Political World: A Schematic Model

Author(s): Pamela Johnston Conover and Stanley Feldman

Pamela Johnston Conover, Stanley Feldman, University of Kentucky

In recent years a substantial metatheoretical shift has taken place in

going beyond the information given (Rumelhart and Ortony, 1977).

Schemas and the Study of Political Belief Systems

Our suggestion that the study of political belief systems be approached

Benefits of the Schema Concept

An application of schema theory does two things that Bennett (1977)

they can be said to have organized, albeit "atomized," ways of looking

processing that are essentially cognitive or nonmotivational in nature.

In applying schema theory to the study of political belief systems,

variability between people in terms of the political schemas that they

Measuring Political Schemas

The empirical study of schemas has lagged behind theorizing in the

As an initial test of our schematic model of political belief systems,

is especially true given that the representativeness of the respondents is

The respondents were also asked about their positions on eleven

Nature of the Political Schemas

Let us begin by considering the nature of the political schemas

The Nature of Political Schemas

IV. Racial beliefs 5.1

A. Religious-liberal sex roles (F51) 41 6

Americans are notorious for their lack of information and well-developed

Structure of the Political Schemas

To this point, we have demonstrated that individuals do have different

8Some caution must be used in interpreting the aschematic percentage. Technically it

In order to do so, however, it is necessary to shift from considering

10 Using correlations between factor loadings as a measure of schematic structure makes

II. Ideological principles

VI. Foreign-affairs beliefs

be hidden or washed out when such differences are ignored. Theoretically,

Schemas and Specific Issue Positions

Govt. Standard Health Income Sch. Preferential Mari

A. Altruistic-positive (F1l) .30

B. Hobbesian-Freudian (F12) -.207

Ill. Economic beliefs

inequities (F32) -.331 -

C. Value of self-reliance (F33) .226

Govt. Standard Health Income Sch. Preferential Mari

B. Liberal sex roles (F5(

dimensions of social beliefs, the two social beliefs schemas overlap

In this paper we have both developed a schematic model to explain

Manuscript submitted 25 October 1982

+ 2. Competition leads to better performance and a desire for excellence.

You might also like