o/
T^^"
A
NONREACTIVE MEASURE OF INFERENTIAL BELIEFS
by
Frank
R.
Kardes
M.I.T. Working Paper No.
May,
1889-87
1987
mr
PJiWJIH
THE MARKETING CENTER
Massachusetts
Institute of
Alfred P. Sloan School of
Technology
Management
50 Memorial Drive
Cambridge, Massachusetts 02139
(617)253-6615
A NONREACTTVE MEASURE OF INFERENTIAL BELIEFS
by
Frank
R.
Kardes
M.I.T. Working Paper No.
May,
1987
1889-87
,
Oct 2
m?
I
A NONREACTIVE MEASURE OF INFERENTIAL BELIEFS
Abstract
Approaches toward measuring inferential beliefs about products are
reviewed and
a
nonreactive measurement procedure
is
advanced.
This
procedure is illustrated empirically in a study demonstrating that the
cognitive operations involved in deducing inferential beliefs from
set of product claims are automatic and inescapable.
a
Implications for
advertising research and for public policy are discussed.
In recent years,
concrete,
there has also been a decrease in the use of
factual claims in advertising and a corresponding increase in
the use of abstract claims that are open to multiple interpretations
(Shimp,
1979;
Shimp and Preston,
1981).
Consequently, the problem of
identifying misleading advertising has shifted from determining if
a
particular claim is factually grounded to determining whether or not a
specific claim leads consumers to draw inferences about product
performance that do not correspond to actual product performance.
The primary focus of this article is on the measurement of
consumer inferences.
An inference is defined as a conclusion that is
derived from a set of premises on the basis of
a
rule, principle or
template that associates the premises to the conclusion in
subjectively logical fashion (Hastie,
Hence,
1983).
a
inference
formation involves going beyond directly observable events in
accordance with the rules of induction (Nisbett and Ross,
1974; Wyer and Carlston,
Wyer and Carlston,
1979) or deduction (McGuire,
1980; Wyer,
1960; Wyer,
1974;
1979).'
Measuring Inferential Beliefs:
A
Review
of the Marketing Literature
Although many consumer researchers agree that inference processes
are Important (Bettman,
Johnson and Levin,
1979;
1985;
1978; Olson and Dover,
Dover,
Meyer,
1978),
1981;
Huber and McCann,
1982;
Mizerski et al.,
few empirical
inference processes have been conducted.
1979;
1982;
Olson,
investigations of consumer
The reason for this is that
methodological difficulties may have discouraged researchers from
studying this topic.
How should inferences be measured?
Further,
consumers are unable to access their inferential processes through
introspection (Nisbett and Wilson 1977), what process -tracing
1
if
methodologies are available to researchers?
Five approaches toward measuring inferential beliefs have been
proposed by marketing researchers, the paraphrase, normative-belief,
salient-belief, comparative-belief, and psychometric procedures.
paraphrase procedure (Preston,
1967;
Preston and Scharbach,
The
1971)
involves exposing subjects to advertising claims and providing them
with five test-items for each claim.
Two of the test-items are
accurate paraphrases (according to the experimenter's definition of
accuracy), and the critical test-item is the illogical item.
As an
example of an illogical item, consider the Bell ad that stated "You'll
get a royal welcome when you call Long Distance" (If X then Y)
.
One
illogical conversion of this claim is "if you don't call Long Distance
you won't get a royal welcome" (If
X'
the Y'
)
.
If subjects judge this
illogical item to be an accurate reflection of the stated or implied
content of the original claim, the claim
is
considered misleading.
The normative-belief procedure (Dyer and Kuehl,
1975)
Gardner,
1978;
involves measuring consumers' normative beliefs about the
attributes associated with
a
particular product class.
are compared to measures of consumers'
associated with
that brand.
a
These measures
beliefs about the attributes
specific brand, after exposure to an ad pertaining to
Discrepancies between normative product class beliefs and
specific brand beliefs suggest that the target ad
is
potentially
mis leading.
The salient-belief procedure (Armstrong et al.,
involves
1979)
asking consumers to judge the validity of various test-items that are
implied by claims presented
in
an ad.
These judgments are weighted by
importance ratings (Wilkie and Pessemier,
operationalized
in
1973), where importance is
terms of the relevance of
2
a
belief for
a
purchase
If consumers'
decision.
differ as
a
salient-belief scores on
a
set of false claims
function of exposure or non-exposure to an ad, the ad is
considered misleading.
The comparative-belief procedure (Russo et al.,
1981)
involves
measuring the beliefs of consumers who read an explicit claim, who did
not read an explicit claim, and who read a corrected (i.e., not
misleading) claim.
If belief in an implied claim is greater in the
explicit-claim group than in the no-claim and corrected-claim groups,
then the explicit claim is deemed misleading.
Schmittlein and Morrison (1983) developed
a
psychometric model
that removes the effect of yeasaying or acquiescence bias (Webb et al.,
1981)
in measuring comprehension.
consumers explicit claims about
a
This procedure involves showing
product.
Immediately afterward, they
are asked to answer some true-false questions pertaining to these
claims.
The correct answer to some of these questions is true and the
correct answer to the remaining questions is false.
of true comprehension for a communication,
The average level
the average level of
yeasaying for a communication, and the level of difficulty for each
question
is
estimated by equating the expected proportion of correct
answers for true and false questions with the observed proportion of
correct answers for true and for false questions,
respectively.
The major limitation common to each of these procedures
is
that
they involve directly asking consumers about the various inferences
that they may have formed.
In
responding to this type of inquiry,
consumers may either search memory for an answer or they may formulate
an answer on the spot
.
Thus,
prompt inference formation
inference
.
,
t he
measurement procedure itself may
rathe r than assessing a pre-existing
For example, suppose that consumers were exposed to an ad
3
stating that Ocean Spray contains more food energy than orange or
tomato juice.
is
^
One false implication of this claim is that Ocean Spray
more nutritious than orange or tomato juice (actually,
If a researcher asks "is Ocean Spray more
higher in calories).
nutritious than orange and tomato juice?" and
is
it's just
a
consumer says it is, it
impossible to determine whether this belief was formed before the
consumer received the researcher's leading question, or if the question
itself induced the consumer to form the false belief.
A Nonreactive Measurement Procedure
Information that is stored in memory
is
derived from two sources:
External sources (e.g., information that was explicitly stated in an
ad)
and internal sources (e.g., self -generated inferences,
Previously acquired knowledge
elaborations, and arguments).
is
often
used to fill in missing details in externally-provided information (see
Taylor and Crocker,
1981) and explicit and inferred information is
often "unitized" (Hayes-Roth,
into a single integrated cognitive
1977)
representation (Bransford and Franks,
1972).
Bransford and Johnson,
This representation is later retrieved in an all-or-none
fashion, and, as a consequence,
in
1971;
individuals often experience difficulty
distinguishing between information that was presented explicitly and
information that was inferred on the basis of the explicitly presented
information.
This type of confusion can be useful
in
determining
whether or not inferential beliefs were formed prior to measurement.
If a consumer forms an
inferential belief while reading a product
claim, he may subsequently be unable to determine whether or not the
inferred piece of information was stated explicitly in the claim.
an inferential belief was not
formed, on the other hand,
this nature is unlikely to occur.
4
If
confusion of
It is
important to note that confusion between explicitly
presented information and information that was inferred on the basis of
the explicitly presented information does not occur under all
Sometimes individuals can accurately discriminate between
conditions.
perceived and inferred information that
is
One
stored in memory.
factor that may influence an individual's ability to determine whether
or not a particular piece of information was inferred is the manner in
which that piece of information was encoded, or stored, in memory.
According to Kahneman (1973), there
is
a
limit on the amount of
capacity that is available for performing various cognitive tasks.
Moreover, some cognitive operations require
considerable amount of
a
this limited capacity for their performance, whereas other operations
require only
These cognitive operations are referred
small amount.
a
to as controlled versus automatic processes,
and Shiffrin,
1977;
Shiffrin and Schneider,
respectively (Schneider
Automatic processes
1977).
can operate in parallel with other tasks and are not influenced by
information load, whereas controlled processes are strongly influenced
by manipulations that divide attention (Bargh,
Schneider,
Hasher and Zacks
1984;
,
1979;
1984;
Fisk and
Schneider and Shiffrin,
1977;
Recent evidence, however, suggests a
Shiffrin and Schneider,
1977).
continuum rather than
qualitative distinction between automatic and
a
controlled processes (Kahneman and Treisman,
1984;
Shiffrin,
in press).
Consider the task of learning how to drive an automobile.
first,
At
the procedure is quite difficult and a great deal of effort is
required (e.g.,
a
controlled process).
With sufficient practice,
however, driving becomes easy and effortless.
process is useful because
behaviors (e.g.,
it
This automatization
enables the individual to perform other
listening to the radio, engaging in
5
a
conversation,
etc.) while driving.
Performing several tasks simultaneously would not
be possible if each of these tasks imposed a large demand on the
limited capacity information processing system.
In an impressive series of studies,
Johnson demonstrated that
accurate discrimination between perceived and inferred information is
likely to occur when the processes mediating the derivation of inferred
information from perceived information are controlled and effortful
(Johnson and Rahe,
1981; Johnson et al.,
1979,
1980,
On the
1981).
other hand, confusion between perceived and inferred information is
likely to occur when the processes mediating the derivation of inferred
information from perceived information are automatic and effortless.
Thus, when a consumer exhibits confusion about the origin of a
piece of information that was not stated explicitly in an ad,
a
researcher may conclude that this information was inferred by the
consumer prior to measurement and that the cognitive mechanisms
underlying this inference process were automatic and effortless.
absence of confusion, however,
is
difficult to interpret.
The
Confusion
may fail to occur because (1) the inference was not formed prior to
measurement, or (2) the inference was formed prior to measurement via a
controlled and effortful process.
The Present Study
In the present study,
presented to subjects.
several sets of product claims were
Each set consisted of two premises of the form
"A implies B" and "B implies C" (Loken,
In addition,
a
1981;
Loken and Wyer,
1983).
conclusion of the form "A implies C" was presented
explicitly In half of the sets and omitted from the remaining sets.
To Illustrate, subjects were asked to rate the validity of several
product claims of the form "Stresstabs contain
6
B
vitamins" (A implies
B)
and "B vitamins give you energy" (B implies C).
On the basis of
these premises, subjects may infer the missing conclusion that
"Stresstabs give you energy" (A implies C).
A recognition-memory technique was employed to determine whether
or not subjects drew this inference.
This technique involved asking
subjects to rate the validity of several sets of propositions, and
subsequently asking them to complete a surprise recognition
questionnaire.
Some of the test-conclusions in this questionnaire were
identical to the target-conclusions considered earlier (e.g., targets)
and some were foils (i.e., conclusions that had not been presented
earlier).
Subjects were asked to indicate whether or not they
recognized each conclusion and they were also asked to judge how
confident they were that each conclusion either had or had not been
presented earlier.
Of critical concern, then, was the recognition
confidence associated with test-conclusions that had not been presented
earlier.
Correct Reco g nition of Conclusi ons That Had Been Presented
It was
reasoned that correct recognition of
a
conclusion that
actually had been presented earlier (a "hit") may reflect
That is, an explicitly
straightforward direct retrieval process.
presented conclusion may be encoded in
fashion when it is first considered.
a
a
relatively unadulterated
Later, during
a
recognition task,
this cognitive representation may be retrieved directly.
One factor that may influence correct recognition is prior
knowledge.
Considerable evidence exists showing that prior knowledge
facilitates the learning of new information and enhances memory for
this information (Bettman 1979, Chase and Simon 1973, Johnson and Russo
1984, Taylor and Crocker 1981,
Voss et al.
7
1980).
Thus,
recognition
accuracy may be greater for conclusions pertaining to familiar than to
unfamiliar products.
False Recognition of Conclusions That Had Not Been Presented
False recognition of
a
conclusion that had not been presented
earlier (a "false alarm"), on the other hand,
implies that this
conclusion was inferred from previously considered premises, and that
the inferred and presented propositions were integrated into
coherent cognitive unit.
a
single,
Later, during a recognition task, the entire
cognitive unit is retrieved from memory and the subject experiences
difficulty in distinguishing between explicit and inferred information.
This type of confusion indicates that the inference was formed prior to
measurement and that automatic cognitive operations mediated the
formation of this inference.
One factor that may influence false recognition is the relatedness
of the propositions that imply the missing conclusion.
Deductive
principles may guide inference processes only when the premises are
logically related.
To test this prediction, control conditions were
constructed for the purpose of comparing false recognition responses to
a
conclusion when logically related premises were presented (e.g., "A
implies B" and "B implies C") to false recognition responses to a
conclusion when logically unrelated premises were presented (e.g.,
implies D" and "E implies C"
)
.
These premises were presented in
A
a
random order, with the constraint that no two premises from the same
set may follow each other.
}Ience,
each premise was surrounded by
premises from other sets and conclusions sometimes occurred prior to
their relevant premises.
This procedure was employed to minimize
pressures on the subject to engage in deductive reasoning.
To summarize, a
2
(presented or omitted conclusion) X
8
2
(familiar
or unfamiliar product) X
2
(logically related or unrelated premises)
factorial design was employed.
dependent measure.
It was
Recognition confidence served as the
predicted (1) that recognition confidence
for targets (i.e., presented conclusions) may be greater for
conclusions pertaining to familiar products than for conclusions
pertaining to unfamiliar products, and (2) that recognition confidence
for foils (i.e., omitted conclusions) may be greater for conclusions
that are supported by logically related premises than for conclusions
that are associated with logically unrelated premises.
Method
Overview
Respondents were 36 undergraduates attending
university who volunteered to participate in
opinions."
a
a
large midwestern
study about "beliefs and
Respondents first judged the subjective likelihood with
which each of 44 product claims was either true or false.
Next they
received a set of math problems and were instructed to solve them as
quickly and as accurately as possible.
This task was employed to
prevent rehearsal and eliminate short-term memory effects on the final
task,
a
surprise recognition confidence task.
The Belief Questionnaire
Examples of the product claims contained in the belief
questionnaire are presented in Table
1.
The belief questionnaire
consisted of 16 two-statement sets of logically related premises (e.g.,
"A implies D" and "E implies C").
Half of these sets contained a
conclusion (e.g., "A implies C") and half did not.
Further, half of
the sets pertained to familiar products (e.g., Nytol, Exedrin, Geritol,
Rolaids, Stresstabs, No Doz, Alka Seltzer, and Tylenol) and half
pertained to unfamiliar products (e.g., Ijemphorin, Allogen, Adrinyl,
9
.
Seratol, Myacin, Cedatol, Tolarin, and Gelutal).
Postexperimental
interviews revealed that subjects did not suspect that these unfamiliar
products were fictional.
Insert Table
1
about here
Two sets of syllogisms were presented in each cell of the
2
experimental design.
In addition,
2
X 2 X
four conclusions with no premises
were included as filler product claims.
These claims pertained to
familiar and unfamiliar products not mentioned in the other sets.
Subjects were asked to judge the validity of each claim on
a
scale
ranging from -4 (Extremely unlikely to be true) to +4 (Extremely likely
to be true)
It should be noted that subjects were asked to judge the product
claims without being explicitly asked to learn them or think about them
in relation to one another.
Moreover, the claims were presented in
a
random order, and no two claims belonging to the same set occurred
together.
Hence, subjects experienced no experiment-induced pressure
to engage in deductive reasoning or to consider conclusions that had
not been presented.
The Recognition Confidence Task
Immediately after completing the belief questionnaire, subjects
solved math problems for 10 minutes.
term memory effects.
This was done to eliminate short-
Next they received
a
surprise recognition task in
which they were asked to rate the confidence with which they believed
they either had or had not seen each of 16 test-conclusions.
Half of
these statements had been presented earlier, and half had not.
Subjects were asked to indicate how confident they were that they
10
,
either had or had not seen each statement earlier in the session on a
scale ranging from -4 ("l am extremely certain
statement") to +4 ("l am extremely certain
T
I
have not seen the
have seen the statement").
Results
A
2
(presented or omitted conclusion) X
product) X
2
(familiar or unfamiliar
(logically related or unrelated premises) repeated
2
measures analysis of variance was performed on recognition confidence
This analysis indicated that
ratings of the conclusion statements.
recognition confidence was greater when the conclusion had been
presented than when
it
had not been presented, F(
familiar than when
.001, when the product was
35) = 6.68,
p <
.02,
,
35) = 196.17,
g <
was not familiar, F(
1
,
35) = 6.28, p <
significant presence X familiarity, F(
presence X relatedness, F(
1
1
,
35) = 7.15,
,
.02.
Moreover,
35) = 4.65, p <
p < .02,
.04,
and
interactions were
The familiarity X relatedness interaction (F = 1.95) and the
three-way interaction (F
<
1)
were not significant.
Correct Recognition
Recognition confidence ratings of presented conclusions as
a
function of product type and premise type are presented in the first
and second columns of Table
2.
A 2
(familiarity) X
2
(relatedness)
repeated measures analysis of variance performed on these ratings
yielded a familiarity effect,
F(l,
35) = 8.54,
As predicted,
p
<
1
and when the premises were logically related than
when they were not related, F(
found.
it
1
.01,
and no other main pffects or interactions.
recognition confidpnce was greater for conclusions
pertaining to familiar products (M = 1.83) than for conclusions
pertaining to unfamiliar products (M = .86).
11
Insert Table
2
about here
False Recognition
False recognition confidence ratings as
a
function of product type
and premise type are presented in the third and fourth columns of Table
2.
A 2 (familiarity) X 2
(
relatedness) repeated measures analysis of
variance performed on these ratings revealed a relatedness effect, F(l,
35) = 16.43, p,
<
.001,
and no other main effects or interactions.
As
hypothesized, recognition confidence was greater for conclusions that
were implied by logically related premises (M, = -1.95) than for
conclusions that were associated with logically unrelated premises (M =
-2.92).
This finding suggests that when logically related product
claims are presented, consumers can construe the implications of these
claims automatically and effortlessly.
Discussion
Together, the findings indicate that recognition confidence
ratings can be useful for extrapolating consumers'
inferential beliefs
without directly measuring or influencing these beliefs.
The present
nonreactive measurement procedure enables researchers to measure
consumers'
inferential beliefs without prompting consumers to engage in
inferential reasoning.
Instead,
inferential beliefs that were formed
spontaneously prior to measurement are assessed.
The results are also consistent with the "reality monitoring"
model of memory (Johnson and Raye,
1981) which focuses on the processes
by which people determine whether a given piece of
information
retrieved from memory originated from an external source (e.g., it was
actually perceived) or an internal source (e.g.,
12
it
was inferred or
People experience little difficulty in determining
self-generated).
the origin of a piece of information when it was inferred through an
effortful process.
When an inference process is not effortful, on the
other hand, the perceived vs.
inferred discrimination cannot be made
confidently.
Some advertisers have taken advantage of consumers'
inability to
distinguish between perceived and inferred information by designing
messages that do not require consumers to engage in effortful
inferential reasoning processes (Preston,
1977; Wilkie et al.,
1984).
Instead, many advertising communications are designed to prompt
consumers to draw inferences automatically and effortlessly.
An automatic process is activated "whenever a given set of
regardless of a subject's
external initiating stimuli are presented,
attempt to ignore or bypass the distraction
1981, p.
117).
(Shiffrin and Dumais,
Moreover, an automatic process does not require
conscious awareness or intention and the operation of an automatic
process cannot be prevented or suppressed.
automatic process is its inescapability
.
Thus,
the key feature of an
Misleading ads that lead
consumers to draw invalid inferences automatically, then, are
particularly insidious, because consumers are defenseless against such
practices.
The present nonreactive measurement procedure offers a
means for identifying advertising claims that prompt consumers to form
invalid inferences automatically.
This procedure possesses the
advantage of enabling the researcher to determine if
inference had been formed before the consumer
measures or queries.
Further,
is
a
particular
exposed to any
the present procedure may prove to be a
useful technique for addressing empirically difficult but important
issues pertaining to how consumers generate inferences about products.
13
Footnotes
1.
Wyer (1974; Wyer and Carlston,
1979) has described the
principles of deduction quantitatively:
where
Pb
is true;
premise A
refers to the subjective probability with which conclusion
P^
is
and
P^^i
B
refer to the subjective probabilities with which
true and not true,
respectively; and
Pb
/
a
and Pb
/
a'
refer
to the conditional subjective probabilities with which B is true if A
is
true and is not true, respectively.
A great deal of empirical
support for this model has been accumulated (for reviews see Wyer,
1974; Wyer and Carlston,
1979).
Other logical inference principles include Kelley
s
(1973)
covariation, augmentation, and discounting principles; Anderson's
(1981) information integration principles;
principles (Abelson,
2.
1981;
Hastie,
1981;
and various schematic
Taylor and Crocker,
1981).
For details on the Ocean Spray case and on other misleading
advertising cases handled by the Federal Trade Commission, see Preston
(1977) and Wilkie et al.
(1984).
14
References
Psychological status of the script concept.
Abelson, R.P. (1981).
American Psychologist 36 715-729.
,
,
Foundations of Information Integration Theory
Anderson, N.H. (1981).
New York: Academic Press.
,
Detecting and
Armstrong, G.M., Gurol, M.N., and Russ, F.A. (1978).
Journal of Consumer
correcting deceptive Advertising.
Research 6, 237-246.
,
Automatic and conscious processing of social
Bargh, J. A. (1984).
information.
In R.S. Wyer and T.K. Srull (Eds.), Handbook of
Hillsdale, NJ:
social cognition (Vol. 3, pp. 1-43).
Lawrence Erlbaum.
An information processing theory of consumer
Bettman, J.R. (1979).
Addison-Wesley.
choice Reading, MA:
.
The abstraction of
Bransford, J.D., and Franks, J.J. (1971).
linguistic ideas. Cognitive Psychology 2, 331-350.
,
Contextual prerequisites
Bransford, J.D., and Johnson, M.K. (1972).
for understanding: some investigations of comprehension and
recall. Journal of Verbal Learning and Verbal Behavior
11
717-726.
,
Chase, W.G.,
and Simon, H.A.
4, 55-81.
Psychology
(1973).
Perception
in chess.
,
Cognitive
,
Dover, P. A. (1982).
Inferential belief formation: an overlooked
concept in information processing research.
In A. A. Mitchell
(Ed.), Advances in consumer research (Vol. 9, pp. 187-189).
Ann Arbor, MI:
Association for Consumer Research.
Memory as a function of
Fisk, A.D., and Schneider, W. (1984).
attention, level of processing, and automatization.
Journal
of Experimental Psychology:
Learning, Memory, and Cognition
10, 181-197.
Gardner, D.M. (1975).
Deception in advertising;
approach. Journal of Marketing 39, 40-46.
A
,
conceptual
,
L.
and Zacks, R.T. (1979).
Automatic and effortful processes
Journal of Experimental Psychology:
General
in memory.
356-388.
108
Hasher,
,
,
,
Hastie, R. (1981).
Schematic principles in human memory.
In E.T.
Higgins, C.P. Herman, and M.P. Zanna, (Eds.), Social
cognition:
The Ontario symposium (Vol. 1, pp. 39-88).
Hillsdale, NJ:
Lawrence Erlbaum.
Hastie, R. (1983).
Social inference.
In M.R. Rosenzweig and L.W.
Porter (Eds.), Annual review of psycholog y (Vol. 34, pp. 51115
542).
Palo Alto, CA: Annual Reviews Inc.
Evolution of cognitive structures and
Hayes-Roth, B. (1977).
Psychological Review 84 260-278.
processes.
,
,
The impact of inferential beliefs on
Journal of Marketing Research 19 324-
Huber, J., and J. McCann.
product evaluations.
(1982).
,
,
333.
Johnson, E.J., and Russo, J.E.
learning new information.
542-550.
Product familiarity and
Journal of Consumer Research 11
(1984).
,
Reality monitoring.
Johnson, M.K., and Rahe, C.L. (1981).
Psychological Review 88 67-85.
,
,
,
Reality monitoring:
Johnson, M.K., Rahe, C.L., and Durso, F.T. (1980).
Second perceptions and thoughts.
Bulletin of the Psychonomic
Society 15, 402-404.
,
Johnson, M.K., Rahe, C.L., Foley, H.J., and Foley, M.A. (1981).
Cognitive operations and decision bias in reality monitoring.
American Journal o f Psychology 94 37-64.
,
,
Johnson, M.K., Rahe, C.L., Wang, A.Y., and Taylor, T.H. (1979).
Fact
and fantasy:
The roles of accuracy and variability in
confusing imaginations with perceptual experiences.
Journal
of Experimental Psychology: Human Learning and Memory 5,
229-240.
,
Johnson, R.D., and Levin, l.P. (1985).
More than meets the eye:
effect of missing information on purchase evaluations.
Journal of Consumer Research 12 169-177.
,
Kahenman, D. (1973).
Prentice-Hall.
The
,
Attention and effort
.
Englewood Cliffs, NJ:
Kahneman, D.
and Treisman, A. (1984).
Changing views of attention and
automaticity
In R. Parasuraman and R. Davies (Eds.), Varieties
of attention (pp. 29-61).
New York:
Academic Press.
,
.
Kelley, H.H. (1973).
The processes of causal attribution.
Psychologist 28, 107-128.
American
,
In A.
B. (1981).
Memory for logically-deduced conclusions.
Mitchell (Ed.), Advances in consumer research (Vol. 9, pp.
348-353).
Association for Consumer Research.
Ann Arbor, MI:
Loken,
A.
B., and Wyer, R.S. (1983).
Effects of reporting beliefs in
syllogistically related propositions on the recognition of
unmentioned propositions.
Journal of Personality and Social
Psychology 45, 306-322.
Loken,
,
Lynch, J.G.,
and Srull, T.K.
(1982).
16
Memory and nttentional factors in
Concepts and research methods.
consumer choice:
Consumer Research 9, 18-36.
Journal of
,
A syllogistic analysis of cognitive
McGuire, W.J. (1960).
In M.J. Rosenberg, C.I. Hovland, W.J.
relationships.
McGuire, R.P. Abelson, and J.W. Brehm (Eds.) Attitude
New Haven, CT: Yale
organization and change (pp. 65-111).
University Press.
A model of multiattribute judgments under
(1981).
uncertainty and informational constraints.
Journal of
M arketing Research 18, 428-441.
Meyer, R.J.
,
and Kernan, J.B. (1979).
Mizerski, R.W.
Golden, L.L.
The attribution
Journal of Consumer
process in consumer decision making.
Research 6, 123-140.
,
,
,
and Ross, L. (1980).
Human inference:
Nisbett, R.
Strategies and
Englewood Cliffs, NJ:
shortcomings of social judgment
Prentice Hall.
,
.
Telling more than we can know:
Nisbett, R.E., and Wilson, T.D. (1977).
Psychological Review
Verbal reports on mental processes.
84, 231-259.
,
Inferential belief formation in the cue
(1978).
utilization process.
In H.K. Hunt (Ed.), Advances in
Ann Arbor:
consumer research (Vol 5, pp. 706-713).
Association for Consumer Research.
Olson, J.C.
,
Olson, J.C. and Dover, P. A. (1978).
Cognitive effects of deceptive
advertising.
Journal of Marketing Research 15 29-38.
,
,
I.L. (1967).
Logic and illogic in the advertising process.
Journalism Quarterly 44, 231-239.
Preston,
,
Preston, I.L. (1977).
The FTC's handling of puffery and other selling
claims made 'by implication'.
Journal of Business Research
5, 155-181.
,
Preston, I.L., and Scharbach, S.E. (1971).
Advertising:
meets the eye? Journal of Advertising Research 11
,
More than
,
19-24.
Russo, J.E., Metcalf, B.L., and Stephens, D. (1981).
Identifying
misleading advertising.
Journal of Consumer Research 8,
119-131.
,
Schmittlein, D.C., and Morrison, D.G. (1983).
Measuring
miscomprehension for televised communications using truefalse questions.
Journal of Consumer Research j.0, 147-156.
,
Schneider, W.
and Shiffrin, R.M. (1977).
Controlled and automatic
human information processing:
I. Detection, search, and
attention.
Psychological Review 84 1-66.
,
,
17
,
'
In R.C. Atkinson, R.J.
Attention.
Shiffrin, R.M. (in press).
Herrnstein, G. Lindzey, and R.D. Luce (Eds.), Stevens
New York:
handbook of experimental psychology (2nd ed.).
Wiley.
The development of
Shiffrin, R.M., and Dumais, S.T. (1981).
In J.R. Anderson (Ed.) Cognitive skills and
automatism.
Hillsdale, NJ: Lawrence
their acquistion (pp. 111-140).
Er Ibaum.
Controlled and automatic
Shiffrin, R.M., and Schneider, W. (1977).
Perceptual learning,
II.
human information processing:
Psychological
automatic attending, and a general theory.
Review 84, 127-190.
,
Social-psychological (mis )representations in
(1979).
28television advertising.
Journal of Consumer Affairs
13
Shimp, T.A.
,
,
40.
and Preston, I.L. (1981).
Deceptive and nondeceptive
consequences of evaluative advertising.
Journal of
Marketing 45, 22-32.
Shimp, T.A.
,
Taylor, S.E. and Crocker, J. (1981).
Schematic bases of social
information processing.
In E.T. Higgins, C.P. Herman and
M.P. Zanna (Eds.), Social cognition:
The Ontario symposium
(Vol 1, pp. 89-134). Hillsdale, NJ: Lawrence Erlbaum.
and Spilich, G.J. (1980).
Voss, J.F., Vesonder, G.T.
Text generation
and recall by high-knowledge and low-knowledge individuals.
Journal of Verbal Learning and Verbal Behavior 19 651-667.
,
,
,
Webb, E.J., Campbell, D.T., Schwartz, R.D., Sechrest, L.
and Grove,
J.B. (1981).
Nonreactive measures in the social sciences
Boston:
Houghton Mifflin.
,
.
Wilkie, W.L., McNeill, D.L., andMazis, N.B. (1984).
Marketing's
scarlet letter'
The theory and practice of corrective
advertising.
Journal of Marketing 48 11-31.
:
.
,
Wilkie, W.L. and Pessemier, E.A. (1973).
Issues in marketing's use of
multi-attribute attitude models.
Journal of Marketing
Research jm, 428-441.
,
Cognitive organization and change:
(1974).
An information
processing approach
Potomac, MD:
Lawrence Erlbaum.
Wyer, R.S.
.
Wyer. R.S. and Carlston. D.E. (1979).
Social cognition,
attribution.
Hillsdale, NJ:
Lawrence Erlbaum.
inference, and
Table
1
Examples of the Belief Statement Sets
Logically Related Propositions
A Implies B:
B
implies
C:
A implies C:
Taking Excedrin will
relieve tension headaches.
Medication that relieves
tension headaches may cause
stomach upset.
Taking Excedrin may cause
stomach upset.
Logically Unrelated Propositions
A implies D:
E
implies
A
implies C:
C:
Seratol helps relieve
Insomnia.
Intestinal disorders often
cause a loss of appetite.
Seratol often causes
of appetite.
19
a
loss
Table
2
Recognition Confidence for Conclusions as
a
Function
of Product Familiarity and Type of Premises
Omitted
Conclusions
Presented
Conclusions
Types of
premises
Related
Unrelated
Note.
Familiar
products
Unfamiliar
products
Familiar
products
1.67
1.99
1.06
0.65
-1.93
-2.74
Unfamiliar
products
-1.96
-3.10
Higher numbers indicate greater recognition confidence.
2°
t;.5?^f:-
03
1
1
Date Due
Lib-26-G7
mil Hill
3
11
nun I mini
nTinniniiiitiiiuifiiiiii
TDflD DDM
-]
33
533