establishing predictors of latitude of · pdf filelatitude of acceptance-rejection and...
TRANSCRIPT
05~
SPECIAL REPORTS 421
ESTABLISHING PREDICTORS OFLATITUDE OF ACCEPTANCE-REJECTION AND
ATTITUDINAL INTENSITY:A CONIPARISON OF ASSUMPTIONS OF
SOCIAL JUDGMENT ANDAUTHORITARIAN PERSONALITY THEORIES
JAMES C. McCROSKEY and MICHAEL BURGOO~
T "VO different theoretical positionsabout the nature of attitudes and
attitude change have provided the basisfor a weal th of recent research. Social
Judgment Theoryl claims that peoplehave latitudes of acceptance-rejectionand attitudinal intensities which are
source and/or concept-specific. That is,people have varying degrees of accept-ance-rejection of sources and conceptsthat depend on who the source is andwhat the topic of communication hap-
. pens to be. According to this theory,people have wide latitudes of accept-ance or rejection on certain specific is-sues and much narrower latitudes on
others. The intensity of prime attitud~
inal position is also claimed to be topic-
dependent. Many studies have been con-ducted based on this assumption with-
out empirically establishing its validity.
The typical experimental paradigm usesone attitudinal issue and proceeds to testderivations of Social Judgment Theory
without comparisons across topics.
Mr. McCroskey is Professor and Chairman inthe Department of Speech Communication atWest Virginia University. Mr. Burgoon is As-sociate Profesor of Speech at the University ofFlorida.
1 See Carolyn W. Sherif. Muzafer Sherif,and Roger E. Nebergall, Attitude and AttitudeChange (Philadelphia: W. B. Saunders Com-pany. 1965): and Carolyn W. Sherif and Mu-zafer Sherif (eds.), Attitude. Ego-lnvoluemmt,and Change (New York: John Wiley and Sons.Inc.. 1968) pp. 105-139.
SPEECH MO.\'OGRAPHS. T'ol. n. .Vvv. 19i.J
Rokeach2 in his development of theoret-
ical propositions about the authoritarianpersonality posits what appears to be anopposite assumption. He claims thatpeople evaluate sources and conceptswithout regard to topic. This theory sug-
gests that people have an enduring per-sonality syndrome that predicts howopen or accepting they are of conceptsand people in general. This focuses onglobal classifications for persons andnot on specific topics and sources. Anas yet untested deductive speculationfrom Rokeach's work would suggest that
highly dogmatic people should havewider latitudes of rejection on all topics.Moreover, the highly dogmatic personwould evaluate sources in a similar
closed manner. In general, low dogmatic
types should be more open-mindedacross topics and sources and have nar-rower latitudes of rejection.
In his original research, Rokeach
sought to establish the dogmatism scaleto be free of political ideology. Theearlier F-Scalea measured authoritarian-
ism of the right. The dogmatism scalesought to measure authoritarianismacross the political spectrum. The politi-cal liberal is supposedly as capable of
2 Milton Rokeach. The Open and ClosedMind (New York: Basic Books, 1960).
3 T. W. Adorno. Else/Frenkel. Brunswick,Daniel J. Levinson. and R. Nevitt Sanford. TheAuthoritarian Personality (New York: Harperand Row. 1950).
422 SPEECH MONOGRAPHS
being closed-minded as his counterparton the right. Thus, a liberal would havea wide latitude of rejection on conserva-tively worded topics or conservativesources; the dogmatic conservative wouldhave wide rejection latitudes on liberaltopics and sources.
The purpose of this investigation wasto test these opposing conceptions. Thequestion of whether latitudes and primeattitudinal positions are topic-specific orare person-specific and best predicted bya personality syndrome was the primeconcern in this study. A second im-portant step was to determine if the lib-
eral-conservative nature of the topic orsource predicted either how wide the
latitudes were or whether dogmatism wasa predictor of evaluation of messagesand sources across the political spectrum.
METHOD
Subjects were 98 undergraduates atMichigan State University. Using theDiab method,4 subjects reacted to 10topics on six seven-interval semanticdifferential-type scales. The tOpics variedon political liberalism-conservatism. On
each topic, latitude of acceptance, lati-tude of rejection, and attitude scores
were computed for each subject. The at-titude score had a potential range of 6(least favorable) to 42 (most favorable).
In addition, 10 sources were selected
for evaluation. On an a-priori basis, fivesources were labeled conservative
(George Wallace, Melvin Laird, RonaldReagan, Barry Goldwater, and RichardDaly). The five other sources were called
-4See for example Lufty N. Diab, "SomeLimitations of Existing Scales in Measurementof Social Attitudes," Psychological Reports, 17(1965), 427-430, and Lufty N. Diab, "Studies inSocial Attitudes: 1. Variations in Latitudes ofAcceptance and Rejection as a Function ofVarying Positions on a Controversial SocialIssue." Journal of Social Psychology, 67 (1965)283-295. This measures both prime attitudeposition and determines mean latitude of accep.tance and rejection scores.
liberal (Eugene McCarthy, Edward Ken-nedy, Nelson Rockefeller, George Mc-Govern, and William Fulbright). Thesubjects evaluated the sources on sixseven-interval semantic differential-typescales again using the Diab technique toestablish latitudes and prime attitudinalpositions.
The 40 item dogmatism scale was ad-
ministered to all subjects and each per-son was asked to rate himself on twoliberal-conservative semantic differen-
tial-type scales. Scores on this measurecould range from 2 (most liberal) to 14(most conservative).
RESULTS
The first statistical analysis was per-formed on the latitude of acceptance,latitude of rejection, and prime attitudescores. If Social Judgment Theory is cor-rect and the latitude and attitude scores
are topic-specific, two different factOrstructures might appear. The first 'al-ternative would be to have no imerpre-table results because each individual
would idiosyncratically mark the scales
and randomly vary across concepts andsources. A second possibility would beto have each source or concept emergeas a factor. That is, 10 concept factorswould emerge with the latitude of ac-ceptance and rejection and attitude in-tensity loading on the latent factor.
If the evaluations are person-specific,the latitude of acceptance scores for all10 topics should load on one prime fac-tor. Similarly the 10 latitude of rejectionscores should form a second factor; the
attitude intensity scores would emergeas a third factor.
The data were submitted to a varimax
rotation to produce factors. Unities wereinserted in the diagonals. Table 1 in-cludes the rotated three factor solution
for the concepts.The factor loadings on latitude of ac-
SPECIAL REPORTS 423
TABLE 1ROTATED FACfoR STRUcruRE . TEN CONCEPTS
The Expected Factor Loading is based on Rokeach's Authoritarian Personality. This aSSUIDI?Sperson-specific evaluation. ' ' ' " :"', , '
.Highest loading."Only one item did not load on the expected factor. Variance accounted for by three factors.53%. " " ,',.
ceptance and reJection are strong andrelatively pure. The 10 acceptance scoresacross all topics form one factor; the 10rejection scores fonn, another, factor.This conforms to expectations that thelatitudes are common to people and arenot topic-specific. The attitudinal in-tensity factor is less strong and pure butindependent of the Iatitude scores. In
other words, peoples' attitudinal polaritydoes not influence widths of latitude
of acceptance and/or rejection.
A similar' factor anal ysis was per-formed on the' two latitude scores and
attitudinal ~osition scores on the 10sources. ,The results are presented inT bl 9' .. :' , ,a e -. '"
This factor structure in Table 2 is an
almost perfect confirmation of the per-son-specific prediction. All five a-prioriliberal sources loaded on one factor;moreover, the a-priori conservativesources loaded on an independent factor.One source on the liberal (Fulbright)
FACTOR 1 FACTOR 2 FACTOR 3 EXPECTEDLatitude of Latirude of Prime FACTOR
Concept. Measure Acceptance Rejection Attitude
Acc .59. .28 .13 1Rej .33 '.60+ .18 2Att .16 .21 .46. 3
2Acc .74. .25 " .03 1Rej .14 .81. .01 2Att .13 .24 ,48. 3
3Acc .48. .46 .06 1Rej .50 .55+ .01 2Att .10 .10 .73. 3
4Acc .82+ .09 .18 1Rej .04 .8se .Il 2Att .35 .19 .39+ 3
5Acc .86. .15 .05 1Rej .10 .61. .12 2Att .03 .08 .59+ 3
6Ace .70. .40 .03 1Rej .48 .548 .08 2Au .38. .24 .35 3-.
7Ace .80. .01 .01 1Rej .01 .778 .14 2Au .09 .13 ' , .45. 3
8Ace .84. .09 .03 1Rej .06 .788 .Il 2Att .27 .27 .40+ 3
9Ace .89- .03 .03 1Rej .20 .76+
"
.01 2..Att .12 .04 .648 3
10Acc .818 .10 .12 1Rej .02 .768 .00 2Au .26 ',' .14 I' .718 3
and one on the conservative (Laird) hadless strong loadings. This is probablybecause these were the least well knownsources and this induced variance in the
evaluations. The loadings on acceptanceand rejection are extremely high andpure. Obviously across all sources, peo-ple responded similarly on the latitudemeasures; this would seriously call intoquestion the notion that the latitudes
are source or topic dependent.
The preceding factor analyses merelyconfirm that people tend to demonstrate
regularities in width of latitude of ac-ceptance and rejection and attitudinal
intensity across sources and topics. Thenext step in this investigation was to de-termine if dogmatism predicted latitudes
and attitudinal intensity. Using multi-ple regression techniques to covary theliberal-conservative measure and pre-dict from dogmatism, dogmatism pre-dicted only attitudes toward conservative
topics. High dogmatics also tended to
be more intensely unfavorable toward
liberal sources (Beta Weight = .19, P
424 SPEECH MONOGRAPHS
TABLE 2
ROTATEDFACTORSTRUCTURE- TEN SOURCES
FACTOR t FACTOR 2 FACTOR 3 FACTOR 4Attitude- AttitUde-
LatitUde of Latitude of Liberal Conservative ExpectedSource Measure Acceptance Rejectin Source Source Factor
Richard DalyAcc .79+ .13 .06 .08 IRej .10 .80+ .09 .15 2Att .11 .07 .29 .63+ 4
Melvin LairdAcc .90+ .07 .08 .03 IRej .05 .84+ .04 .03 2Att .14 .08 .36 4-+ 4. :>
George WallaceAcc .87+ .15 .03 .07 IRej .25 .80+ .01 .02 2Att .09 .03 .17 .64+ 4
Ronald ReaganAcc .88+ .16 .05 .03 IRej .17 .82+ .05 .13 2Att .02 .03 .03 .68+ 4
Barry GoldwaterAcc .92+ .10 .03 .06 IRej .17 .84+ .05 .13 2Att .08 .17 .17 .69+ 4
George McGovernAcc .88+ .03 .10 .06 IRej .03 .84+ .04 .06 2Att .17 .10 .60+ .13 3
Nelson RockefellerAcc .8S+ .16 .01 .01 IRej .OS 8+ .12 .04 2. IAtt .04 .07 .77+ .09 3
Eugene McCarthyAcc .87+ .22 .09 .10 1Rej .23 .84+ .Il .08 2Au .15 .04 .80+ .12 3
William FulbrightAcc .87+ .15 .02 .11 IRej .09 .86+ .03 .05 2Att .22 .04 .40. .23 3
Edward KennedyAcc .81+ .29 .04 .02 1Rej .20 .75+ .10 .09 2Att .16 .03 .58. .15 3
+Highest loading. All items loaded on expected factOr. Variance accounted for by four factors:660/0.
SPECIAL REPORTS 425
<.06). Dogmatism did not predict thesize of latitude of acceptance or rejec-tion on either source or topic.
The self-reports of liberalism-conserv-atism also tended to be a significant pre-dictor. When dogmatism was covaried,significant prediction obtained on atti-tudes toward concepts and sources; how-ever, this measure did not predict thewidth of the latitUdes except in one in-stance.
TABLE 3MULTIPLE REGRESSION OF
LIBERALISM-CONSERVATISM ON LATITUDES OF
ACCEPTANCE REJECTION AND AlTITUDESTOWARD SOURCES AND CONCEPTS
PREDICTOR; Liberalism -Conserva tism
A negative beta weight indicates a negativecorrelation with conservatism.
The results indicate that conserva-tives are more favorable toward conserv-
ative concepts (Beta Weight = .37, P<.01) and sources (Beta Weight = .45,P <.01). Liberals were more favorabletoward liberal concepts (Beta Weight =.21, P <.04) but no significant differenceswere found in their evaluations of lib-
eral sources. Only in one instance did theself-report measure latitude. Conserva-tives have a wider latitude of acceptanceon conservative' topics (Beta Weight =.25, P <.02) but there are no other sig-nificant latitUde predictions. There is noindication that conservatives or liberals
differ on latitudes of rejection; more-over, liberals do not have wider latitude
of acceptance on liberal topics.
DISCUSSION
The results question the original as-
sumption posited by Social JudgmentTheory that latitudes of acceptance-re-jection and attitude intensity are topic-specific. The factor analyses clearly indi-cate that people have relatively invari-ant widths of latitudes of acceptance and
rejection across topics and sources.Moreover, absolute intensity of attitudi-
nal position does not appear to be topic-specific but remains a constant markingbehavior across topics and sources. Peo-
ple who have wide latitUdes have themon all topics; conversely people with .narrow latitudes exhibit this marking
behavior across topics and sources.Those people who deviate a great dealfrom neutrality, whether positive or neg-ative, polarize their judgments onsources and topics in general. This find-ing is important for it can allow re-searchers to specify variables that are
person-specific to further test SocialJudgment Theory assumptions.
The second part of this investigationlooked at two person-specific variablesto attempt to predict attitude intensity,and latitude of acceptance-rejection.Only limited success in predicting fromdogmatism scores obtained. High dog-matics were more favorable toward con-
servative topics and low dogmatics weremore favorable toward liberal topics.However, no evidence demonstrated that
dogmatism affected latitudes of accept-
CRITERION BETA WEIGHT P
Attitudes Toward -.21 <.04Liberal Concepts
Latitudes of Acceptance .1O NSDLiberal Concepts
Latitudes of Rejection .10 NSDLiberal Concepts
Attitudes Toward .37 <.01Conservative Concepts
Latitudes of Acceptance .25 <.02Conservative Concepts
Latitudes of Rejection -.04 NSDConservative Concepts
Attitudes Toward -.17 <.08Liberal Sources
Latitudes of Acceptance .Il NSDLiberal Sources
Latitudes of Rejection .10 NSDLiberal Sources
Attitudes Toward .45 <.01Conservative Sources
Latitudes of Acceptance .06 NEDConservative Sources
Latitudes of Rejection -.16 NSDConservative' Sources
/"
426 SPEECH MOl,'OGRAPHS
ance or rejection. These findings indi-cate that more research needs to be com-
pleted to determine if dogmatism isindependent of political ideology. Giventhe findings that high dogmatics aremore favorable on conservative topicsand low dogmatics are more favorable
about liberal topics, this assumptionmust be questioned.5 It would also seemthat high dogmatics should demonstratedifferences in openness or acceptanceand closed-mindedness or rejectionwhen compared with low dogmatics. The
latitude measures do not support this.In conclusion, dogmatism alone is not
5 Similar results questioning the constructvalidity of the dogmatism scale as a measure ofgeneral authoritarianism have obtained. In factlow dogmatism has been associated with leftismand high dogmatism with rightism. See, forexample, Herbert W. Simons and Nancy NeffBerkowitz, "Rokeach's Dogmatism Scale andLeftist Bias," SM, 36 (1969), 459-463, and MiltonRokeach, "Political and Religious Dogmatism:An Alterna tive to the Authoritarian Person-ality," Psychological MonograPhs, 70 (No. 18,whole no. 425, 1956), and Herbert W. Simons,"Dogmatism Scales and Leftist Bias," SM, 35(1968), 149-153. The leftist bias found in thisinvestigation is consistent with these priorresearch findings.
useful predictor and asresearch, a liberal bias
a particularlywith previousobtained.
The self-report liberal-conservativemeasure did, as expected, indicate thatconservatives were more favorable to-
ward conservative topics and sources;liberals were more favorable toward lib-
eral topics, but not liberal sources. How-ever, this measure did not predict lati-tudes except in the case of conservativeshaving wider latitudes of acceptance onconservative topics. This measure cer-tainly has limited utility in understand-ing the person-specific variables that in-fluence the latitudes of acceptance-'rejection. It does, however, have predic-tive utility with attitudinal position.
With all of the empirical support forSocial Judgment Theory available, fur-
ther research needs to search for predic-tors of latitUdes of acceptance and re-
jection. The concepts of latitude of ac-
ceptance-rejection are central to thetheory and their determinants need to
be empirically established.