Content-Type: text/html

The Role of Media Examples in
The Heuristic Process Model of Cultivation Effects
 
 
 
Rick W. Busselle, Ph.D.
 
 
 
 
 
 
 
 
Washington State University
Murrow School of Communication
Pullman, Washington 99163-2520
Phone: (509) 335-6838
e-mail:  <a href="/cgi-bin/wa?LOGON=A3%3Dind9812D%26L%3DAEJMC%26E%3D7BIT%26P%3D1492360%26B%3D--%26T%3DTEXT%252FPLAIN%3B%2520charset%3DUS-ASCII" >[log in to unmask]</a>
 
 
 
 
 
 
 
The Role of Media Examples in
The Heuristic Process Model of Cultivation Effects
 
 
ABSTRACT
 
 
        This study explores the influence of specific examples on  judgments.  Analysis
is carried out in light of The Heuristic Process Model of Cultivation Effects
(Shrum, 1996).
Ss (197) completed a traditional cultivation survey, and six weeks later were
divided into two experimental conditions.  Condition 1 performed an exemplar
accessibility task measuring the amount of time required to think of an example
of an extra-marital affair, a shooting, and an African-American doctor.  They
then completed a social judgment survey asking about the prevalence of those
constructs in society. Condition 2 respondents performed the two tasks in
reverse order.
        Respondents in Condition 2 provided examples marginally more quickly than those
in Condition 1.  This indicates specific examples (or closely related
constructs) are involved in heuristically processed social judgments.
        This study also explored the sources of respondents' most accessible examples.
It found fictional examples of shootings and affairs came predominantly from
theater movies, pay-cable, and video rental, not traditional television sources.
 
The Role of Media Examples in
The Heuristic Process Model of Cultivation Effects
 
        After more than three decades and 300 publications of cultivation1 research
(Morgan & Shannahan, 1996) investigators have identified a cognitive process
that at least partially explains the relationship between television exposure
and social reality judgments.  Shrum, et al. refer to this mechanism as "the
heuristic process model of cultivation effects" - referred to heretofore as
HPMCE (Shrum, 1997b; Shrum, Wyer, & O'Guinn, 1998).  The model is especially
important because other theoretical explanations for the cultivation
relationship have received little empirical support (Hawkins & Pingree; Mares,
1996; Shapiro, 1991; Shrum, 1995).  Conversely, Shrum, et al. have published a
series of studies presenting data consistent with their model (O'Guinn & Shrum,
1997; Shrum, Wyer, & O'Guinn, 1998; Shrum, 1996).  The HPMCE establishes that
heuristic processing is partly, possibly largely, responsible for the
relationship between viewing and social judgment.  However, more research is
needed to fully understand the nature of that heuristic process.  The present
article investigates specifically the role of examples and exemplar
accessibility2 in light of the HPMCE.
          An availability heuristic (Tversky & Kahneman, 1973) is at the heart of
HPMCE. When individuals make social judgments or estimates, such as those
required in a cultivation survey, they often do not systematically consider and
evaluate all of the information they possess related to the judgments at hand.
Instead, because of lack of motivation, distraction, or time constraints,
individuals may use a processing strategy that requires less cognitive effort
(for a summary of heuristic processing, see Fiske & Taylor, 1991).  When
employing an availability heuristic individuals base judgments solely or partly
on the ease with which information related to the judgment comes to mind.  For
example, when asked what percent of marriages end in divorce, a person who
easily can think of a divorced friend is more likely to estimate divorce is more
prevalent than someone who has difficulty recalling a similar example.
According to the HPMCE, "greater frequency and more recency of viewing will
cause instances of such things as crime and violence to be more accessible in
memory for heavy viewers than for light viewers" (Shrum & O'Guinn, 1993, p. 483;
also see, Shrum, 1997, Shrum, Wyer & O'Guinn, 1998).
        In HPMCE research accessibility has been operationalized as the amount of
time required for respondents to answer a social judgment question. Using this
measure of information accessibility the researchers have produced convincing
empirical evidence.  Shrum and O'Guinn (1993) found heavier viewers judged the
world to be more violent, war to be more deadly, illegal drug use to be more
prevalent, and prostitution more patronized than lighter viewers.  More
importantly, heavier viewers took less time to make those judgments than lighter
viewers.  Another study (Shrum, 1996) found judgment speed mediated the
relationship between amount of soap opera viewing and judgments about the
prevalence of crime and the occupational prevalence of doctors and lawyers.
O'Guinn and Shrum (1997) found heavy soap opera viewers made estimates about
affluence more quickly than lighter viewers.  Again, those who responded more
quickly estimated higher percentages.  Finally, Shrum, Wyer, and O'Guinn (1998)
indicate data to be published demonstrate that forcing respondents to process
systematically reduces or eliminates the relationship between television viewing
and social judgment.  Reportedly, when heuristic processing was prevented the
cultivation effect was eliminated.
        The empirical evidence extant builds a convincing argument that information or
accessibility underlies the cultivation relationship.  However, thus far little
is known about form of the information that is accessed when heuristic
processing is used in cultivation judgments.
        The accessibility of several different types of information has the potential
to influence judgments (Sherman & Corty, 1984).  The ease with which individuals
are able to imagine an event or the ease with which they can retrieve a prior
judgment has been shown to influence subsequent social estimates (Shrum, 1996,
pp. 49-50; also see, Manis, Shedler, Jonides, Nelson, 1993).  In the literature
regarding judgments about the traits of individuals (e.g. whether a person is
honest or dishonest), there is evidence that both the accessibility of specific
examples and the accessibility of abstractions or prototypes influence judgments
heuristically (Sherman & Klein, 1994; also see Fiske & Taylor, 1991, p.392).
Specific Portrayals and Social Judgment
        Cultivation theory indicates overall television consumption is responsible for
the construction of social reality (Gerbner, Gross, Morgan, & Signorielli,
1994).  However, others have pointed out the importance of specific portrayals.
For example, Greenberg (1988) argued, "critical images may contribute more to
impression-formation and image-building than does the sheer frequency of
television characters and behaviors that are viewed" (p. 100).  Research has
supported this perspective empirically by demonstrating the influence of
specific portrayals on social judgments immediately following and several days
after exposure to experimental stimuli (Gibson & Zillmann, 1994; Tamborinin,
Zillmann, & Bryant, 1984).
        Tamborini, et al. (1984) found exposure to a film portraying unjustified
violence and exposure to a crime documentary lead subjects to higher estimates
they would be victimized by violence and greater concern about violence in urban
areas as compared to the estimates of subjects who watched a control film.
Further, the effect of the documentary film stimulus lasted for three days while
the effects of the fictional stimulus did not.  To demonstrate that the severity
of an example influences subsequent social judgments Gibson and Zillmann (1994)
manipulated the severity of one specific example contained in a newspaper story.
They found even though the actual rate of carjacking incidents was given in each
stimulus story, subjects exposed to more severe examples contained within the
stimulus stories estimated the incidence of carjackings to be higher.  This
effect was evident both immediately after and one week after exposure (Gibson &
Zillmann, 1994).
        In both sets of findings describe immediately above a heuristic process is one
explanation.  Explaining their findings, Tamborini, et al. argued exposure to
the fictional violence "primed schemata relevant to crime and victimization that
temporarily biased perceptions" (p. 509).  The researchers did not specify, nor
did the data indicate, what type of relevant constructs may have been primed.
Similarly, HPMCE research has provided no direct indication of the type of
information that is more accessible among heavy than light viewers.
        This study attempts to link social judgments with specific examples or
exemplars3 through a heuristic process.  The rationale is this:  If specific
examples are involved when individuals make social judgments, those examples
should be more accessible after a social judgment is made than before.  For
instance, if one is asked to estimate the percent of Americans who have
extra-marital affairs, and if one bases her estimate on the accessibility of an
example of an affair, that exemple should be more accessible after the estimate
than before.  This logic prompts the following hypothesis.
H1: Judgment related examples will be more accessible after judgments are made
than before.
 
        The HPMCE proposes heavier television viewers have more accessible examples of
constructs featured often on television.  If exemplar accessibility underlies
the HPMCE process the following hypothesis is consistent with the model.
H2: For content areas portrayed often in the media, heavy viewers will report
more examples from television than light viewers.
 
        Previous research also makes it possible to hypothesize about some
characteristics of examples that should make them relatively more or less
accessible (Higgins & King, 1981).  Some of these characteristics are especially
relevant to media content.
Perceived Realism
        The extent to which individuals perceive media content as representing the real
world has been shown to vary on as many as six dimensions (for reviews, see
Busselle, 1995; Potter, 1988).  Survey research has demonstrated that the
perceived realism of television in general, rather than specific programs or
portrayals, may account for more variance in social judgment than television
exposure itself (Potter, 1986).
        Experimental research has demonstrated that the verisimilitude of specific
pieces of information impacts on that information's ability to influence
judgment.  Mares (1996) showed subjects a twenty minute stimulus tape containing
both ostensibly "real information," contained in a TV newscast and "fictional
information," contained in an advertisement for a movie.  She found a positive
relationship between mistaking fictional information as real and subsequent
social judgments related to that information.  She also found the cultivation
relationship was mediated by the number of times subjects mistakenly identified
fictional information as real: a higher number of fiction to real mistakes was
associated with higher cultivation estimates while a higher number of real to
fiction mistakes was associated with lower cultivation estimates.  One
explanation for these findings is that information perceived as real is more
accessible in memory, thereby having a greater opportunity to influence
judgment.
        The cognitive processing literature indicates perceived realism of information
does indeed influences its accessibility (Gerrig & Prentice, 1991; Johnson,
Hashtroudi, & Lindsay, 1993; Potts, St. Johns & Kirson, 1989).  Potts, St. Johns
and Kirson (1989) found individuals compartmentalize information they believe to
be fictional separately from real-world knowledge.  They demonstrated this by
having subjects read information that was ostensibly real or fictional.  They
then primed half of each group to think about either the experimental
information or unrelated, real-world information.  Subjects from the fictional
condition answered questions about the experimental story faster when primed to
think about the experimental facts.  But subjects in the real information
condition answered the same questions faster when primed to think about
unrelated, real-world facts, supporting the compartmentalization argument.
        If individuals recall specific examples when making social judgments, those
examples may be real or fictional and may come from the media or the real world.
If examples are observed in the media they may be perceived as more or less
representative
of reality. This leads to the following hypotheses.
        H3: Real exemplars will be more accessible than fictional exemplars.
and
H4: Fictional exemplars perceived as more real will be more accessible than
those perceived as less real.
 
        As described previously, Gibson & Zillmann (1994) found the severity of a
provided stimulus example influenced subsequent judgments related to that
example.  This effect should also occur when examples are recalled from real
life experience for the purpose of making social judgments.  Thus;
        H5: More severe examples will be more accessible
 
 
METHOD
 
Sample and Design.  Students (463) from introductory telecommunication courses
at a large Midwestern university completed an initial survey containing measures
of television exposure, impulsiveness, need for cognition, demographics, and
other scales (perceived realism of television, involvement with television, and
social desirability).  Six to eight weeks later 211 of the initial respondents
volunteered for a second ostensibly unrelated experiment.  In the second
procedure equal numbers of males and females were randomly assigned to one of
two conditions.  In condition 1 respondents completed an exemplar accessibility
task and then a social judgment survey on a computer.  Condition 2 respondents
completed the two tasks in reverse order -- the social judgment survey before
the accessibility task.  Fourteen respondents were dropped prior to data
analysis; five did not understand the task or did not take it seriously.  Nine
self-identified as having been raised outside the U.S. This left 197 respondents
(107 male, 90 female).  Ages ranged from 18 to 32 (mean = 20.3 years).
Social Judgment Items.  Three social judgment questions designed to tap
constructs frequently portrayed on television were imbedded among 20 similar
items related to a broad range of topics (See Exhibit 1).  Respondents answered
each question by entering a two-digit response on the numbers pad of a computer
keyboard.  Respondents answered five practice questions on the computer before
beginning the task.
_____________________________________________________________________
 
Exhibit 1
 
        Social Judgment Items
 
        1. "What percent of married Americans have extra-marital affairs?"
        2. "What percent of medical doctors are African-American?"
        3. "Other than in hunting accidents and war, what percent of Americans are
             shot with a gun in their lifetime?"
 
        Exemplar Category Prompts
 
        1. "A HUSBAND OR WIFE HAVING AN AFFAIR"
        2. "AN AFRICAN-AMERICAN MEDICAL DOCTOR"
        3. "SOMEONE SHOT WITH A GUN"
_____________________________________________________________________
 
 
        Exemplar Accessibility.  Three category prompts were created to match the
social judgment questions described above (Exhibit 1). Each category appeared
separately on the computer screen.  Categories were rotated so that each was
presented first, second, and third and equal number of times.  Respondents were
instructed to press a key as soon as they thought of a specific example of the
category on the screen.  They were instructed that their example could come from
any source, including personal observation, another person, or the media.  After
pressing the key indicating they had thought of an example they described their
example and answered as series of questions about it.  Exemplar accessibility
was operationalized as the amount of time (in centiseconds) that elapsed between
the appearance of the prompt on the screen and the respondent pressing a key to
indicate they had thought of an example.  Data were collected on a IBM personal
computer using a version of MELLAB (1994) software.  Respondents practiced on
two media related categories before performing the task.  They were asked to
think of "an actor or actress" and "two cars in a high speed chase."
                Example Source.  Respondents were asked where they observed each example they
reported.  Sources were categorized as personal observation, interpersonal
source, one of twelve media sources, and an other category.
        Perceived Realism and Seriousness of Exemplar. The researcher read all exemplar
characteristic items to each respondent.  Response options were one-to-ten
scales (See Exhibit 2). A four item scale was used to measure respondents
perceived realism of fictional examples.  These were selected from nine items.
Principal axis factor analysis placed these four items on one factor, with
factor loadings of .70 or greater.  The four items averaged to form a scale
called Exemplar Realism (Mean = 5.8, Alpha = .89).
        Exemplar Seriousness was measured with two items (see Exhibit 2).  These two
items were averaged (mean = 7.69).  Correlation between them was .82.  The
seriousness item was not used with the 'African-American doctor' category.
 
____________________________________________________________________
 
Exhibit 2
 
Perceived Realism Items
 
1.      Events like the one you told me about do happen in real life.
                [People like the person you told me about do exist in real life.]*
 
2.      Incidents [or people] like the one you told me about are quite  common in
real life.
 
3.      The event [person] you told me about could happen [exist]       in real life.
 
4.      Events like the one you told me about happen in real    life ALL OF THE TIME.
                [People like the person you told me about ARE VERY COMMON.]
 
Seriousness Items
 
1. Compared to most events [people] like this, how serious is the example you
gave me?
 
2. Compared to most events [people] like this, how severe is the example you
gave me?
 
                *               Words, phrases, and sentences in brackets were used for examples of Police
Officers or African-            American Doctors.
        **      Responses could range from 1 to 10, where 1 = strongly disagree and 10 =
strongly agree.
______________________________________________________________________
 
 
        Television Measures.  Respondents were asked how many hours per week they
watched programs in each of 11 categories (e.g. sitcoms, soap operas, crime
dramas, etc.).  Each category was accompanied by at least two program examples.
The 11 categories were summed to create a total television exposure measure
(mean = 12.3 hours, range =  0 to 39 hours).
 
RESULTS
        Before hypothesis tests are reported there is a description of the types of
examples and their sources generated by respondents when asked to think of their
first example from  the three categories.
        Examples Reported.  Real and fictional examples were distributed quite
similarly in both conditions.  Respondents reported more fictional than real
examples in all categories (Table 1).  In condition 1 respondents reported 39
real and 50 fictional examples of extra-marital affairs; 26 real and 64
fictional examples of African-American doctors; and 26 real and 62 fictional
examples or shootings.  In condition 2 they reported 42 real and 50 fictional
examples of extra-marital affairs; 22 real and 70 fictional examples of
African-American doctors; and 29 real and 60 fictional examples of shootings.
_____________________________________________________________________
 
Table 1: Real and Fictional Examples in Conditions 1 and 2
 
        Condition 1     Condition 2
 
Example Real    Fiction Total   Real    Fiction Total
 
African-American M.D.   26              64      70      22              70      92
 
Affair  39              50      89      42              50      92
 
Shooting        26              62      88      29              60      89
 
 
_____________________________________________________________________
 
        Example Sources.  The vast majority of real examples of extra-marital affairs
originated from personal (14) and interpersonal sources (60).  The difference
between these two categories is subtle and may not be important.  Those who
reported personal observation of an affair typically said they learned a friend,
relative, or parent was romantically involved with someone other than their
spouse, and the respondent had observed them together.  Interpersonal sources
typically were similar situations, but the respondent had not observed the pair
together personally.  In the vast majority of cases (74 of 81 times)
respondents' first example of a real affair came from interpersonal sources.
Conversely, examples of fictional affairs came from the media, and mostly from
films, rather than television.  Fifty-seven (of 100) fictional examples of
affairs came from movies seen in theaters or on television.  Thirty-one came
from television programs.  Of those, 15 came from night-time drama programs;
only nine came from afternoon soap operas.
        The majority (39 of 48) of real examples of African-American doctors originated
in personal experience.  The majority of fictional examples came from television
dramas (82 of 134) and situation comedies (36 of 134).  The most commonly
identified African-American doctor named as a first example was the Dr. Benton
character from the TV program ER.  The second most popular fictional example was
the Cliff Huxtable character from the TV show Cosby.  Third most common was the
cartoon doctor on the TV show, The Simpsons.
        Interpersonal experience accounted for the majority of real examples of
shootings (26 of 55).  Most of these respondents knew of someone directly or
indirectly who had been shot with a gun.  In one of the shooting examples
personally observed, a female was shot by her brother with a BB gun.  In the
other four cases the respondent was neither a victim or assailant.  Rather she
or he was at a party where someone was shot, observed the aftermath of a
shooting, or heard shots fired but did not see the event.  Fictional examples of
shootings originated overwhelmingly in movies seen at theaters (50 of 122) and
in rented movies or on pay cable television (43 of 122).  Only 26 (of 122)
fictional examples originated from television programs.
 
______________________________________________________________________
Table 2: Sources of Real and Fictional Examples
 
        Affair  Afr.-Am. Doctor Shooting
 
Source  Real    Fiction Real    Fiction Real    Fiction
        Personal Experience     14      0       39      0       5       0
        Interpersonal   60      1       1       1       26      0
 
        Local News      1       0       0       0       9       0
        National News   0       0       3       0       1       0
        Reality / Talk  4       0       1       0       5       0
 
        Newspaper / Magazine    1       0       1       0       1       0
 
        TV Drama Program        0       15      0       82      0       17
        Situation Comedy        0       4       0       36      0       3
        Soap Opera      0       9       0       5       0       1
        Other TV        0       4       0       2       0       3
 
        Movie / Theater 0       13      0       3       1       50
        Movie / VCR Pay Cable   0       36      0       2       1       43
        Movie / Basic TV        0       8       0       1       0       2
 
        Novel   0       2       1       2       0       1
        Other   1       7       2       0       6       2
 
        Totals  81      100     48      134     55      122
________________________________________________________________________________
__
 
Hypothesis Tests
Outliers. Two types of outliers were of concern in this study.  First, extremely
quick exemplar response times were likely to be produced when respondents
"jumped the gun" and pressed the response key before they had an example in
mind.  Second, extremely slow response times were likely to be produced when
respondents did not attend fully to the task or when they evaluated several
examples before reporting one.  In order to minimize these sources of error, the
quickest and slowest 10-percent of responses were eliminated from each
comparison group (Cameron & Frieske, 1994, pp. 159-160).  For example, the
fastest and slowest 10-percent of response times were removed from examples of
African-American doctors in condition 1 and the fastest and slowest 10-percent
of response times were removed from African-American doctor examples in
condition 2 for comparisons between conditions.  This procedure was carried out
for each comparison analysis.
 
Hypothesis 1: Judgment related examples will be more accessible after judgments
are made than before.
 
 
Respondents in Condition 2 - where the exemplar accessibility task followed the
social judgment questionnaire - reported examples from all three categories more
quickly than respondents in Condition 1.  However, only the difference in
accessibility of African-American doctors was statistically significant.  In
condition 1 respondents took an average of 1.06 seconds longer to report
examples of African-American doctors (t = 2.502, df = 145, p < .02); 1.06
seconds longer to report examples of extra-marital affairs (t = 1.351, df = 141,
ns); and .37 seconds longer to report examples of shootings (t = .591, df = 138,
ns).  These results offer tentative support of Hypothesis 1.   In no case were
judgment related examples less accessible after judgments were made.
_____________________________________________________________________
 
Table 3: Accessibility of Examples in Conditions 1 and 2
 
Example Condition 1 (n) Condition 2 (n)
 
Afr.-Am. Doctor 562(73) 456(74) 2.502   145     .013
 
Affair  824(71) 718(72) 1.351   141     .179
 
Shooting        649(69) 612(71) .591    138     .556
_____________________________________________________________________
 
 
 
 
Hypothesis 2: For content areas portrayed often in the media, heavy viewers will
report more examples from television than light viewers.
 
        Hypothesis 2 was tested by comparing the frequency of fictional to real
examples among light and heavy television viewers using a median (11 hours)
split (Table 4).  It was also tested by comparing the frequency of examples from
different sources reported by heavy and light viewers (Table 4b).  Respondents
reporting watching less than 11 hours of television a week were defined as light
viewers.
        Though a trend in the correct direction of a few percentage points was
observed, there were no statistically significant differences in the
distributions of real and fictional examples between heavy and light viewers for
any of the three examples (Table 4).
The second comparison looked at the sources of  examples, which were divided
into interpersonal, television, film, and other media sources (Table 4b).  When
compared to light viewers, there was a tendency for heavy television viewers to
report more examples from television and fewer from real life or movies.
However, this difference was not statistically significant.  Therefore data
provide no empirical support for Hypothesis 2.
 
 
_____________________________________________________________________
 
Table 4: Percent of Real and Fictional Examples in Conditions 1 and 2
 
        Light Viewers   Heavy Viewers
 
Example Real (%)        Fictional (%)   Real (%)        Fictional (%)
 
Affair  40 (46%)        47 (54%)                41 (44%)                53 (56%)
 
African-American M.D.   25 (28)         63 (72)         23 (24)         71 (76)
 
Shooting        29 (33)         59 (67)         26 (29)         63 (71)
 
 
Total   94 (36)         169 (64)                90 (32)         187 (68)
 
 
Table 4b: Sources of Examples Among Heavy and Light Viewers
 
Source  Light Viewers   Heavy Viewers
 
Personal / Interpersonal        75      (33%)   72      (30%)
 
Television (including TV movies)        97      (42)    120     (50)
 
Film (Theater, VCR, Pay-Cable)  45      (20)    37      (15)
 
Other   14      (6)     12      (5)
_____________________________________________________________________
 
 
 
        Hypothesis 3: Real examples will be more accessible than fictional examples.
Hypothesis 3 was tested by comparing the response times of real and fictional
examples for each of the three example categories.  Again quickest and slowest
10-percent of response times were eliminated from each comparison group.
        Real examples of affairs and shootings were reported more quickly than their
fictional counterparts (Table 5).  However the opposite was true of
African-American doctor examples.  Real examples of affairs were reported an
average of 3.7 seconds faster (t = -3.931, df = 141, p < .001).  Real examples
of shootings were reported an average of 1.04 seconds faster, though the
difference was not statistically significant.  Conversely, real examples of
African-American doctors were reported 2.82 seconds slower than fictional
examples (t = 5.680, df = 145, p< .001).
_____________________________________________________________________
 
Table 5: Accessibility of Real and Fictional Examples
 
Example Real (n)        Fiction (n)
 
Affair  609(64) 979(79) -3.931  141     .000
 
Afr.-Am. Doctor 725(42) 443(105)        5.680   145     .000
 
Shooting        567(45) 671(96) -1.518  139     .131
_____________________________________________________________________
 
 
Hypothesis 4: Fictional examples perceived as more real will be more accessible
than those perceived as less real.
 
        This analysis includes fictional examples only.  Zero-order correlations
between exemplar accessibility and the perceived realism of examples were
computed for all three examples in Condition 1 and Condition 2 separately.  The
results indicate only one statistically significant correlation to support
Hypothesis 4. In condition 2 fictional shootings perceived as more realistic
took less time to access as predicted (r = -.35, n = 48, p < .05).  The
remaining three correlations related to shootings and African-American doctors
also were in the predicted direction but did not reach statistically
significance.  The correlations between perceived realism of affairs and the
accessibility of those measures did not support Hypothesis 4.
_____________________________________________________________________
 
Table 6: Zero-order Correlations between Perceived Realism of Fictional Examples
and Exemplar Accessibility
 
        Condition 1     Condition 2
        Example Accessibility   r (n)   r (n)
 
Affair          .16 (37)                .02 (38)
 
Afr.-Am. Doctor         -.15 (51)               -.08 (58)
 
Shooting                -.08 (48)               -.351(48)
 
 
1 P < .05
_____________________________________________________________________
 
 
Hypothesis 5: More severe examples will be more accessible.
 
        To test Hypothesis 5 correlations between exemplar accessibility and perceived
severity of examples were computed.  A trend was evident in Condition 1
suggesting severity was negatively correlated as expected.  However the
relationships did not reach statistical significance (Affairs; r = -. 18, n =
69, ns; Shootings; r = -.10, n = 59, ns).  In Condition 2 both of these
correlations were near zero (r = -.04 and r = -.03, respectively).
 
DISCUSSION
        The data presented in this study provide mixed evidence regarding the role of
exemplar accessibility in the heuristic processing of cultivation judgments.  As
predicted by the HPMCE, specific examples were slightly more accessible after
social judgments were made than before.  While only one comparison was
statistically significant, the other two were in the predicted direction.  This
suggests either specific examples are accessed in the judgment process or
constructs closely related to these examples are accessed.  For instance, a
prototype to which an example is integral may be accessed priming the specific
example.  Either case is consistent with the HPMCE and refutes questions of
spuriousness in observed relationships between judgment response time and social
judgments.  If faster judgments were due to some third variable, such as
intelligence or impulsiveness, there would be no reason for judgments to
influence exemplar accessibility.
        Examination of the sources of examples is also informative.  The examples
reported by respondents are their most accessible, which should be an indication
of importance to the individual.  If television examples play an important role
in the social judgment process, one would expect a greater proportion of real
examples to originate from news and fictional examples to originate from
entertainment television.  However, a large majority of fictional examples of
both affairs and shootings came from film.  This may reflect different kinds of
viewing experiences and levels of involvement.  While viewers may observe many
examples of violence and relational discord on television, they were more likely
to recall examples from theater, pay-cable or rented video.  If memorable
examples are a reflection of involvement, these media experiences may have a
greater potential to influence social judgments and social perception.  This
observation points to a weakness of the present study: theater attendance,
pay-cable viewing, and video rental were not measured.  Research into the
influences of media, both from a heuristic processing perspective and from other
approaches, should pay particular attention to the role of media fare outside
traditional television programming.
        Further, the public often laments the prevalence of violence on prime-time
television.  However these data warrant speculation that it is not prime-time
and day-time viewing that have the greatest viewer impact.
        The observation that film viewing is responsible for more memorable examples
also has implications for generalizing from laboratory settings to television
exposure.  Given the captive nature and presumably higher involvement extant in
laboratory settings, research in which subjects are shown experimental stimuli
may be more akin to a movie going or video rental experience than to everyday
television grazing.
        It should be emphasized that fictional examples of African-American doctors
originated primarily from television and are counter to many of the points made
above.  There are several possible explanations for this difference.  First,
television characters may be more memorable than television events because
viewers observe characters weekly, but specific events only once. That is, the
same character appears in every episode, but a shooting appears only one time,
excluding re-runs.  Second, these data simply may reflect the content of films.
Popular movies frequently portray shootings and affairs, but rarely portray
African-American doctors.
        One contribution of this study is the employment of naturally occurring and
specific media examples in the study of the social reality construction.  This
methodology has the potential to fill a gap between survey research focused on
general exposure patterns and experimental research using artificial stimuli.
        A second, related contribution is the use of exemplar accessibility
methodology.  This tool may prove useful for investigating the impact of
specific media content.  However, more refinement of this measurement technique
is necessary, as well as more borrowing from cognitive psychology and its
methodologies.
        One important limitation to this study is its use of a college student sample.
A more diverse sample has two obvious advantages.  First generalizations are
limited to college students who are unique in both their viewing levels and
content choices.  Further the lack of variance in viewing amount inherent in
college student samples attenuates many observed relationships.  Future research
should seek a more diverse sample so that statements about observed
relationships can me made with more certainty.
 
 
 
 
 
REFERENCES
 
Busselle, R. W. (1995). A conceptual structure for understanding perceived
realism of media content. Paper presented at the Association for Education in
Journalism and Mass Communication, Washington DC(August 9-12).
 
Cammeron, G. T. & Frieske, D. A. (1994).  The time needed to answer: measurement
of memory response latency.  In A. Lang (Ed.) Measuring Psychological Responses
to Media. Hillsdale, NJ: Earlbaum.
 
Fiske, S. T., & Taylor, S. E. (1991). Social Cognition. New York: McGraw-Hill.
 
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Growing up with
television: The cultivation perspective. In J. Bryant & D. Zillmann (Ed.), Media
Effects: Advances in Theory and Research. Hillsdale, NJ: Earlbaum.
 
Gerrig, R. J., & Prentice, D. A. (1991). The representation of fictional
information. Psychological Science, 2(5), 336-340.
 
Gibson, R., & Zillmann, D. (1994). Exaggerated versus representative
exemplification in news reports. Communication Research, 21(5), 603-624.
 
Greenberg, B. S. (1988). Some uncommon television images and the drench
hypothesis. In S. Oskamp (Ed.), Television As A Social Issue: Applied Social
Psychology Annual (Vol. 8, pp. 88-102). Newbury Park, CA: Sage.
 
Hawkins, R., Pingree, S., & Adler, I. (1987). Searching for cognitive processes
in the cultivation effect. Human Communication Research, 13(4), 553-577.
 
Hawkins, R., & Pingree, S. (1990). Divergent psychological processes in
constructing social reality from mass media content. In N. Signorielli. M.
Morgan (Ed.), Cultivation Analysis: New Directions in Media Effects Research
(pp. 35-50). Newbury Park, CA: Sage.
 
Higgins, E. T., & King, G. (1981). Accessibility of social constructs:
Information-processing consequences of individual and contextual variability. In
N. Cantor & J. F. Kihlstrom (Ed.), Personality, Cognition, and Social
Interaction (pp. 69-121). Hillsdale, NJ: Erlbaum.
 
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source Monitoring.
Psychological Bulletin, 114(1), 3-28.
 
Manis, M., Shedler, J., Jonides, J., & Nelson, T. E. (1993). Availability
heuristic in judgments of set size and frequency of occurrence. Journal of
Personality and Social Psychology, 65(3), 448-458.
 
Mares, M. L. (1996). The role of source confusions in television's cultivation
of social reality. Human Communication Research, 23(2), 278-297.
 
Morgan, M., & Shanahan, M. (1997). Two decades of cultivation research: an
appraisal and meta-analysis. Communication Yearbook, 20, 1-45.
 
O'Guinn, T. C. & Shrum, L. J. (1997). The role of television in the construction
of consumer reality.  Journal of Consumer Research, 23, March, 278-294.
 
Potter, W. J. (1986).  Perceived reality and the cultivation hypothesis.
Journal of Broadcasting & Electronic Media, 30(2) 159-174.
 
Potter, W. J. (1988). Perceived reality in television effects research. Journal
of Broadcasting & Electronic Media, 32(1), 23-41.
 
Potts, G. R., St. Johns, M. F., & Kirson, D. (1989). Incorporating new
information into existing world knowledge. Cognitive Psychology, 21, 303-333.
 
Shapiro, M. A. (1991). Memory and decision processes in the construction of
social reality. Communication Research, 18(1), 3-24.
 
Sherman, S. J., & Corty, E. (1984). Cognitive heuristics. In R. S. Wyer & T. K.
Srull (Ed.), Handbook of Social Cognition (pp. 189-286). Hilldale, NJ: Erlbaum.
 
Sherman, J. W.& Klein, S. B. (1994). Development and representation of
personality impressions.  Journal of Personality and Social Psychology, 67(6)
972-983.
 
Shrum, L. J., & O'Guinn, T. (1993). Processes and effects in the construction of
social reality. Communication Research, 20(3), 436-471.
 
Shrum, L. J. (1996). Psychological processes underlying cultivation effects:
further tests of construct accessibility. Human Communication Research, 22(4),
482-509.
 
Shrum, L. J. (1997). The role of source confusion in cultivation effects may
depend on processing strategy: a response to Mares.  Human Communication
Research 24 (2) 349-358.
 
Shrum, L. J., Wyer, R. S., & O'Guinn, T. C. (1998). The effects of television
consumption on social perceptions: the use of priming procedures to investigate
psychological processes. Journal of Consumer Research, 24(March), 447-458.
 
Shrum, L. J., Wyer, R., & O'Guinn, T. (1998). The effects of television
consumption on social perceptions: the use of priming procedures to investigate
psychological processes.  Journal of Consumer Research, 24, March, 447-458..
 
Smith, E. R. & Zarate, M. A. (1992).  Exemplar-based model of social judgment.
Psychological Review, 99(1), 3-21.
 
Tamborini, R., Zillmann, D., & Bryant, J. (1984). Fear and victimization:
Exposure to television and perceptions of crime and fear. In B. R. Burleson
(Ed.), Communication Yearbook (pp. 492-513). Newbury Park, CA: Sage.
 
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging
frequency and probability. Cognitive Psychology, 5, 207-232.
 
 
 
1 I will follow Shrum's (1995) convention of using the term cultivation to refer
to any investigation of relationships between television exposure and social
judgment.
 
2 Gibson & Zillmann (1994) distinguish between the terms accessibility and
availability.  "Information, although available in memory, is not equally
accessible" (p. 605).  In this paper availability is a dichotomy; information is
either available or not.  Information that is available may be more or less
accessible.  Accessibility is therefore viewed as a continuous variable.
 
3 An exemplar is defined as a cognitive representation of an object, incident or
behavior that is similar to the current target judgment (Smith & Zarate, 1992).
Media Examples and Social Judgment
 
 
 
 
 
25