You are on page 1of 17

Effects of Humor on Persuasion Author(s): Dorothy Markiewicz Source: Sociometry, Vol. 37, No. 3 (Sep., 1974), pp.

407-422 Published by: American Sociological Association Stable URL: http://www.jstor.org/stable/2786391 . Accessed: 24/02/2014 06:41
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to Sociometry.

http://www.jstor.org

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

Sociometry 1974, Vol. 37, No. 3, 407-422

Effectsof Humor on Persuasion


DOROTHY MARKIEWICZ*
Northern Illinois University

on humor'seffects An information-processing analysisis used in examining the attitude changeprocess. Generally: humor integralto or adjacent to a humor's persuasive message does not influencepersuasion significantly; effectson comprehension and source evaluations are inconsistent;and retention does not appear to be altered by humor usage. Severe methodologicalproblems with prior research include inadequate control messages, questionable humor manipulations,inappropriatesettingsfor receipt of humor, limited subject populations, and blatant demand characteristics.Many of these problems plague persuasion research in general. Two theoretical approaches, learning theory and distraction effects, offerrequisiteguidance forfutureinvestigators.

or rhetoric oftensuggest Those concernedwithpolitics,advertising, of a persuasivemessage. that humor enhances the effectiveness Advertising agencies' bias is evidentin that 42 percentof television use some humor (Markiewicz, commercials 1972a,b). Yet, research of humoron messageeffectiveness the effects suggests considering is questionable; goes thus,muchof thisresearch that its contribution effect.This unpublisheddue to the failureto findany significant studiesin orderto as well as unpublished, review considers published, point out serious methodologicalproblemsand steer investigators use of energy. towarda moreexpeditious model of persuasion(Hovland et al. An information-processing 1953) is used to explicate the relationshipof humor to the of a persuasive persuasionprocess. In this approach,the processing actionsof attending includessequentialcognitive communication to, acceptanceof a persuasive accepting,and retaining comprehending,
*The author thanks Anthony Greenwald,Timothy Brock, Thomas Ostrom and advice. Special appreciation and Ladd Wheelerfor theirhelpfulsuggestions is expressedto MartinKaplan and Robert Haccoun fortheircommentson earlier drafts of this review. Requests for reprints should be sent to the author, Northern Illinois University, De Kalb, Illinois 60115. Departmentof Psychology,

407

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

408

SOCIOMETRY

communication. Completionof all stagesyieldspersuasion. Research to determine humor'spersuasive considering impactwillbe reviewed of humoron each of thesestages.Additionally, the effect substantive of the many methodological issues,whichbear on the interpretation will be discussed. Finally, two marginaland conflicting findings, theoretical approachesare suggestedas potentially usefulguidesto future humor-persuasion research. EFFECTS OF HUMOROUS MESSAGES Studiesare consideredunderthreemainsections:(1) studiesthat did not employ controlsforseriousmessages, thereby confounding of humorincorporated intoa message the effects withsuch factors as and sourcevariables; messagearguments (2) studieswhichhave used serious messagesfor control groups; and (3) studies which offer for the effects implications of humor which is externalto (and sometimes unrelated to) the adjacentmessage. SeriousControls Absent Is humor an effective vehicle for persuasion?Relevant studies typicallyemploy satirein producinghumor;those reviewedin this the results sectionomitseriousmessage difficult counterparts making to interpret. That is, the humor-only designsused confound the effect of humor with other variables also likely to influence the serious attitudes. However, these are reviewed to highlight methodological difficultiespresent even in those experiments employing serious control messages. Table 1 summarizes the implications of these studies for humor's effects on the and acceptancestagesof theattitude comprehension changeprocess. of humorous Of the seven studieswhich consideredthe effects threefailedto find messageswithoutcomparablecontrolmessages, any change in attitudesfollowing receiptof the humorousmessage (Gruner,1965, 1966; Zeman, 1967). These three used the same of which the persuasiveness message,a satire attacking censorship, of the humorouscontent) is questionable.The lack of (regardless attitude change in Gruner (1965) may have been due to the pretest-posttest design coupled with instructions givento subjects in "how (Ss): the experimenter (E) indicatedthathe was interested stable or unstable" opinions were (see Papageorgis,1968, for a on persuasion).In the other two reviewof the effectsof warning the Es attempted to reducepossiblemiscomprehension experiments,

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION


TABLE 1 Studies Considering the Effectsof HumorousMessages Reference Comprehension
+
-

409

Results Reten- Attitude tion


+ +

Ratingsof Source

L Humorous Message Only Berlo & Kumata (1956)


Gruner (1965) Gruner (1966) Gruner (1967a)

0 0

Gruner(1971a) Windes(1961) I. Serious Message Control Gibb (1964) Gruner(1967b) Gruner(1970) Gruner(1971b) Kennedy (1970) Kennedy (1972a) Kennedy (1972b) Kilpela (1961) Lampton (1971) Lull (1940) Markiewicz(1972a) Experiment1, Chapter2
Gruner (1972) Zeman (1967)

+ +? 0 + +

+ 0 0
-

0 0 0 0
-

+? 0 0 0 0 +?
0

+ 0 +
0
+
-

Experiment3 McGown (1967) Pokorny(1965) Pokornyand Gruner(1969) Taylor (1964)


Taylor (1972)

Markiewicz (1973)

Experiment2, Chapter2 Markiewicz(1973) Experiment2

Markiewicz (1972b)

0
-

0
-

0 0 0 0

Youngman(1966)

Note: For Humorous Message Only experiments, (in I) "+" indicates a positive effect relative to control condition. "+" indicates a significant difference between humorous and serious messages in II with the more favorable effect following the humorous message (e.g., more persuasion, retention, positive source evaluations due to humorous message). "-" indicates that the more negative effects occurred following humorous than control conditions if in I, or than serious messages in II (e.g., poor comprehension, less retention, less attitude change, lower source evaluations). "O" indicates that no significant differences occurred on the measure comparing humorous message and control conditions (in I) or humorous and serious message conditicns (in II).

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

410

SOGIOMETRY

out itssatiric purpose, yetwereunableto of the messageby pointing changeSs' opinions. due to humorous Of the four studieswhich did observeeffects Gruner only whenSs were told the messages, (1967a) foundeffects of point of the satire-which suggests the possible contribution Windes (1961) did not actually measure demand characteristics. attitudechange,but insteadused as dependentvariablesthe judged of "in terms of campaign effectivenss speechesby Adlai E. Stevenson is equivocal sincethe the candidate'scause." This finding advancing judges might have evaluated the speeches as a functionof their assumed effectivenesswith the electorate, rather than with Effectsobserved in the other two studies (Berlo and themselves. weak, comparedto Kumata, 1956; Gruner,1971a) were generally no-messagecontrol conditions,and not consistenton all of the attitude measures. relevant dependent In the one case (Berloand Kumata,1956) in whichthe source(the with Canadian Broadcasting Company)was evaluated,as contrasted the following were more negative pretestratings, source evaluations satiricmessage. Researchersoften suggestthat Ss are unlikelyto comprehendthe message correctlywithout extra cues about its 1965, 1966, 1967a; Zeman,1969). Thus persuasive purpose(Gruner, humor may interfere with comprehension.Finally, since humor ratingswere generallylacking in these studies,the success of the is not certain, threatens humormanipulation internal whichseriously of the humorvariableis a validity.Inadequate operationalization recurrent problemin thistypeof research. With SeriousMessages Comparisons Studies summarizedin this section focused on whetherthe addition of humor to a message enhances its persuasiveness; these studies have generallyincluded a seriousmessage therefore, control. In addition to those studies measuring attitude change, studies exploringthe effectsof humor on retentionand speaker in thissection. ethosare also considered Humor and attitudechange. Thirteen comparedthe experiments relative persuasive impact of humorous versus serious messages (Gruner,1971b, 1972; Kennedy,1972; Kilpela, 1961; Lull, 1940; Markiewicz, 1972a, 1972b, 1973; McGown, 1967; Pokorny,1965; Pokornyand Gruner, 1969; Youngman,1966). Table 1 summarizes the resultsof these experiments. Ten studiesfailedto findevidence for any difference in persuasiondue to messagehumor;two found tentative of humorous effectiveness messages; supportforthe greater

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

411

and one founda seriousmessage to be morepersuasive. of these studiesemployedone-wayfactorial The majority designs threeconditions:(1) a humorousmessage;(2) a serious containing control message; and (3) a no-messagecontrol. Major dependent or Thurstone. variableswere attitude scales-generallyLikert-type These studieswere characterized by a plethoraof typesof humor, including satire,irony,puns,turnsof phrases, anecdotes,burlesque, hyperbole, understatement, and witticisms.The messages were sometimes developed by professional comedy writers, and sometimes in the typesof humorhad by theresearchers. Variabilities no systematic effecton persuasion.The messages focusedon serious problemssuch as state medicine,censorship(pro or con), goals of higher education, capital punshment,Martha Mitchell, summer school,safety-belt usage,anti-gun of campus legislation, and arming police. Two studiesprovideequivocal evidenceforthe greater persuasive impact of humorousmessages(Gruner, 1971b; Markiewicz, 1972a). In the Gruner study,the comparability of his humorous and serious messages(an Art Hoppe satire versus a shortbiographical sketch from Time, both negativecriticisms of Martha Mitchell)was not discussed.Thus, the morenegative of Mitchell ratings on "character" and "ridiculousness" scales by those reading the satire are uninterpretable (p < .01). In the Markiewicz(1972a) study,on a measure of attitudes,seventh-grade listed-thoughts studentswrote more thoughts agreeing with the humorousmessagethan withthe a morepositive seriousone (p < .04), suggesting attitude towardthe former.However, no significant differences emerged on a more conventionalattitude measure (self-rating scales). Finally,Markiewicz (1973) showedlesspersuasion thehumorous following thanthe serious message (p <.01). This resultmay have been due to the of thehumorous poorercomprehension message(p <.001) or to its sourcetrustworthiness lowerperceived (p < .01). Several authors considered selected variables which might moderatehumor'spersuasiveness. Markiewicz (1972a, 1972b, 1973) foundthat verbalability,source credibility, and initialopinionsdid not consistently moderatethe effect.In her study,however,the source credibility was weak; thus,thisvariablemight manipulation well be addressedin futureresearch.Kennedy (1972) crossedthe humorfactorwith a messagechannel factor(live versusvideotape versus audiotaped speech). Analyses revealed no significant main norinteraction on attitude effect responses. Thus, the pursuitof the elusive increase in persuasiondue to

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

412

SOCIOMETRY

humor has been generally unsuccessful. Humor apparently has no and possible moderatorvariables have simple effecton persuasion, identified. yetto be reliably Humor and retention.Some investigators (e.g., Gibb, 1964; Gruner,1967b; Kennedy, 1970; Taylor, 1964) have hypothesized that humor increases message retention. However, as Table 1 indicates,only one out of elevenstudiessupportedthishypothesis. In this case (Gibb, 1964) a humorousbiologylecturewas retained more than a seriousone. Again,these resultsmustbe significantly interpreted with caution since intact classes were used for each group (i.e., Ss were not randomlyassigned to conditions,and instructor was confounded with condition). In two experiments of the arguments contained (Markiewicz, 1972b, 1973) recognition in the humorousmessagewas significantly inferior to that forthe serious message. The dependentmeasuresin these studies do not and comprehension; distinguish between retention thus, effectson memorycannot be separatedfromthose on initialunderstanding. on messagearguments, shouldbe used to Recognition tests,focusing tap comprehension differences; delayedrecall testsmight be used to index retention.However,in most cases, usinggrossmeasuresof or recognition retention (e.g., Gruner, 1967b, 1970; Kennedy,1970, 1972) no significant differences wereobtained. Humor and source evaluations.Studieswhichconsidervariations of the source as a function in perceptions of humorusage generally on severalsemantic-differential-type employsource ratings scales.In six experiments(Gruner,1967b, 1970; Kennedy, 1972a, 1972b; Markiewicz,1972a, 1973), the humorous source appeared to be morepositively evaluatedthan the serioussource,on one or moreof the following dimensions: "character", qualification(expertise), and dynamism. In severalcases safety(evaluative,trustworthiness), (Lampton,1971; Markiewicz, 1972b; McGown,1967), no significant in source evaluations differences emerged.Finally,severalinvestigators reportedthat the serioussource was evaluatedmore positively than the humorousone (Markiewicz, 1973; Taylor,1972). Thus,the of humoron sourceperceptions effects tendsto be inconsistent, with enhancement moreprobablethanloweredevaluation, suggesting that sourceswho are not highly evaluatedinitially might benefit the from use of humor.This should be testedin a factorial design withsource (high versuslow credibility or liking)and messageappeal (humor versus the typeof humorused is likelyto serious)as factors. Further, influence sourceperceptions. Hostilehumorcould decreaseliking for the source(Gutmanand Priest, 1969).

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

413

Humor and interest ratings. Are humorous messages more interesting than serious ones? Two studies report no significant differences betweenhumorousand seriousmessages (Gruner, 1967b; Lull, 1940), and two suggestthat humorousmessagesare more than comparableseriousones (Markiewicz, interesting 1972a, 1973). Gruner (1970) varied interestingness (high versus low) and humorousness (high versuslow) of a messagein a 2 x 2 factorial design. A dull message was created by rewriting the interesting version to removeelementsof interest (personalwordsand personal sentences).A significant interaction (p <.01) was found,suchthat addinghumorto the dull messageproducedmore favorable interest ratings,as well as source ratings;whereas adding humor to the interesting speech did not. This impliesthat a dull speakerbenefits fromincluding humor.If one is alreadyinteresting, however, littleis to be gainedfromusingit. The Gruner procedure could be repeated, adding an attitudinaldependent measure to determinewhether interestand source evaluationschangeswould be accompaniedby attitude changes. The followingmightbe inferred fromresearchreviewedin this section: There appearsto be no significant difference due to humor at the acceptanceand retention (and possibly comprehension) stages. Humor does appear to increase interest, which impliesa possible increase in attention; however, specific attempts to measure attention are absent. HumorExternalto theMessage The precedingdiscussionreviewedstudiesin which humorwas into persuasive incorporated messages.A major problemwithdirect comparisons of humorous and serious messages is that humor into a messageis likelyalso to affect the interpretation incorporated of the arguments contained.Thus, one is unable to determine what proportion of any observedeffects is due to the generalacceptance createdby the humorand whatproportion is due to the new specific of the arguments interpretation studies (comprehension). Therefore, which consideredthe effectsof humor contiguousto the message, rather than incorporatedinto it, may offer clearer insightinto on the acceptancestageof persuasion. humor'seffects This section reviews studiesthatused contiguous humorwhichwas either relevant or irrelevant to the messages. Relevant contiguous humor. Markiewicz (1972b) varied the inclusion or exclusion of a cartoon in short or longerpersuasive lettersin a 2 x 2 factorial design.A behavioralindex of persuasion

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

414

SOCIOMETRY

who returned enclosed postcards) (i.e., the percentageof recipients main effectof thecartoonand no interaction revealedno significant withmessage length. a In anotherstudy (Markiewicz, 1972b) eithera humorousfilm, a standard usage introduced serious film,or no filmon safety-belt to enforcesafety-belt usage. recordedmessageadvocating legislation was the Subjects weretold that the source of both filmand message same. No significantdifferenceson attitudinal or behavioral intentions measures were found,even thoughthe serioussourcewas than the humorousone (p < rated as significantly betterinformed relevantadjacent humor .008). Thus, the two studies considering in persuasion due to thehumor. foundno increase Irrelevant contiguous humor. In two studies (Smith, 1961; was presented at Festinger and Maccoby,1964), humorousmaterial the same time as Ss were beinginducedto changetheirattitudes. In the Smithstudy,hearinga humorousrecord(Bob Newhart's"The effect on Buttondown Mind Strikes Back") had no significant attitudes towards eating grasshoppers. In the Festinger-Maccoby to a speech study,Ss who watcheda humorousfilmwhilelistening changed their attitudesmore than did those viewingthe speaker. of humorwithdistraction. Both experiments confoundedthe effects That is, any humoreffects may be due to the humor actingas a distractor; any distractor mighthave done as well as the humorous material (see Baron, Baron and Miller, 1973, for a review of and persuasion). distraction Markiewicz(1973) varied the contextsin which five persuasive or messageswereembedded.Messages wereembeddedin a humorous a serious background program. After each persuasive message, opinion data on the focal topic were collected. Humorousversus seriouscontexthad no differential on audienceattitudes. effects of humorexternal the effects to a whichconsiders Thus,research that it does not significantly affectacceptanceof messagesuggests the message. This is consistent with the majorityof the reviewed researchon the persuasiveeffectsof humorincorporated into the message. METHODOLOGICAL PROBLEMS WITH ATTITUDE RESEARCH ON HUMOR EFFECTS Whatis a good controlmessage? fromeach otherbecause of the inclusion or Messageswhichdiffer omissionof humoralso necessarily differ on some otherdimensions.

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

415

For example, many of the studies reviewed insertedadditional humorousmaterialinto a message (Pokorny,1965; Pokornyand Gruner, 1969; Kennedy, 1972a; Gruner, 1967b, 1970). The additional humor made the humorous message longer than the seriousmessage.This might resultin an additional factor(e.g., more in that could have maskeddifferences time to rehearsearguments) to the responses messages.Those studies which tried to control message length had to insert some additional material (e.g., or introduction of new ones). repetition, elaborationof arguments Thus, one cannot determinewhat is responsiblefor differences obtained (or lack of differences)-the manipulation of the independentvariable or some seemingly innocuous aspect of the it is critical forresearchers to use controlmessage.For thesereasons, several issues withinthe same study. If the independent variables yield equivalenteffects across all issues,one can be more confident in attributing theseeffects to the intended rather manipulations than to some confounded factor(s). is funny How funny enough.? of messagesused in prior The variationin degree of funniness researchcomplicatesthe interpretation of results.Sometimesthe humorousmessagesare actuallyonly attempts at humor,a problem commonin studiesusingsatire, a complexand subtleform of humor. Since the degreeof funniness has neverbeen systematically variedin an attitudestudy,it is difficult to determine what effect thisfactor mighthave on responsesto the message. Future researchshould if systematically varythe humorlevel of the messages. Furthermore, the humor is differentially poignantacross studies (and therefore differentially operationalized), then valid comparisons between experiments become precarious. in thelaboratory? How does one measure attention The majorityof the researchdiscussed here was conductedin classroomsor laboratories.The pressure to attend to the messages it probablethatattention was likelyto have been great,making was held constant(high) acrossconditions.Thus, one cannotdetermine fromthese studieswhether in the "real" worldpersons would attend more to a humorousmessagethan to a seriousone. Futurestudies mightmanipulatebase line attentiveness to humorousand serious For example,messagesnot apparently influenceattempts. a part of the experiment and immediate recallcould be mightbe presented, used as an indexof attention.

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

416

SOCIOMETRY

typesof humor? due to different are theeffects What whichcould affect The type of humorused conveysinformation recipients' attitudes towards the source the perceived signifiof arguments.Youngman cance of the issue,and the interpretation of speechescontaining germane therelative effects (1966) considered the point of the communication humor (humor that supplemented or no humor. of it), non-germane an illustration humor, by providing but were foundon measuresof attitudes; differences No significant sinceSs alreadyagreedwiththe thesisof the speech,thiswas a weak test of the persuasiveimpact of the messages. Brandes (1970) of fivecategories of humor, but to testthe persuasiveness attempted (e.g., betweencategories foundthatSs could not tell the difference maynot be as The formalclassification puns versusplays-on-words). as the impliedtone of the humor(cf. Gutmanand Priest, important from effects hostilehumormay have quite different 1969). Critical, slapstick. humorappreciation? inhibit Do serioussettings situations, Since the studiestake place in classroomor laboratory the serious environmentmay well have interferedwith the sampleswho expected a appreciationof the humor. Experimental the serioustask,but insteadreceivedhumormaynot have perceived as funny. McGhee(1971) pointsout thatSs need humorous messages external cues in order to know when it is appropriateto use humor) versusreality "fantasy" assimilation(as when interpreting a serious representation). assimilation (as when interpreting naturalor this problemby creating mightcircumvent Investigators to the presentation more appropriate of humorous familiar settings stimuli. results of humor contaminate How muchdo demandcharacteristics and persuasionstudies? salient Two kindsof demand effects (Orne,1962) are particularly in humor research.First,the "persuasionto be persuaded" effect mighthave induced some Ss to indicatethat theirattitudeswere changed. This problem was especially likely in studies that used 1966), 1965; Gruner, coverstories(e.g., Pokorny, rather transparent and betweenhumorous and could accountforthelack of differences serious conditions.In addition to this demand, Ss in the humor the demandto ratethe message as humorous. conditionsexperienced appearedto be askingthemto ratethe Thus, whentheirinstructors

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

417

obvious attemptat humoras funny, manycomplied.Snyder(1971) has shown that Ss' judgmentsof humor are affectedby demand to determine the characteristics. This makes it particularly difficult More indirectmeasuresof humor success of humormanipulations. evaluations should be used (e.g., see Langevinand Day (1972) fora of physiological of humor). correlates discussion How might variations in thepersuasivemessages affecttheresults of studies? of messages used in thesestudiesadds to the confusion The variety of interpretations. more Complex messagesmake comprehension problematic. Not all studies actually measured comprehension, the althoughauthors often suggestedthat Ss did not understand messages(e.g., Gruner,1965; Zeman, 1967). Messagesalso varyin If the arguments persuasiveness. in a humorous messageare weak,an not be detected.On the otherend extraboost due to humormight cause ceiling that of the continuum, effects powerful messages might would maskadditionalgainsdue to humor. How do subjectpopulationsaffect results? exists about how persons other than Very little information studentsreact to humorousmessages.In additionto the limitation on the generality of the results,the use of studentsas Ss often creates other problems.Students in speech classes hear numerous the persuasion speecheson varioustopics,sometimes including topic, priorto the experiment. Pokornyand Gruner(1969) pointout that this occurred in their experiment.Other studies using speech students might also suffer from this contamination. Second, "evaluation apprehension" (Rosenberg,1965) and compliance to experimental demandcharacteristics are aggravated. Students maybe particularly likely to comply with what they perceiveas the E's wishes,whenthe E is theirteacher.Third,collegesubjectswho have been trainedto value rationalism may reject an emotionalappeal more than would other populations.This is impliedby Rokeach's (1973) findings that those with less formal educttion rank intellectual valueslower. Thus, the use of morevariedpopulations is How does one prove thenullhypothesis? it may be concludedthathumordoes On the basis of thisreview, of a message.The majorityof the not increase the persuasiveness

imperative.

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

418

SOCIOMETRY

studies reviewed, however, suffered in varying degrees from methodological problems that could mask actual differences. humor may not increasepersuasionin even the most Nevertheless, findings sound experiment. These "no difference" methodologically shouldnot be ignored. SOME THEORETICAL IMPLICATIONS to the effects have generally theoriesrelevant ignored Researchers of humor on persuasion.This has resultedin simple approaches with the complexityof the problem.Two theoretical incompatible in applicable to the questionare considered approachespotentially as possiblecompassesto those and are offered lightof priorresearch, thelabyrinth. who enter theory Learning A number of studies have found that associatingan irrelevant (e.g., itspersuasiveness messageaffected witha persuasive reinforcer et Lana, 1965; Rosnow and Rosnow, 1965, 1966; 1965; Janis al., Rosnow and Russell, 1963). If humor functionsas a positive humor placed in temporal proximityto persuasive reinforcer, shouldenhancetheirpersuasiveness. communications The logical research steps should be to show that (1) humor and (2) this learning paradigm) functions as a reward(in a-standard at the affects be directed testing persuasion.Researchmay reward on recipients'agreement effectsof humorpresentation contingent in thatis, usinghumoras a reinforcer, withthe arguments presented; the humorwas an operantsense. In all cases previously investigated, of Ss' responses to the message.If Ss are told to regardless presented theiragreement or disagreement with the message indicateovertly humorpresentedfollowing withthe message agreement arguments, attitudes and in existing recipients' mayproveusefulin strengthening to persuasion. resistance increasing in a classical stimulus, Humor may operate as an unconditioned conditioningsense. That is, if humor elicits a positive affective response,a message paired with humor mighteventuallyelicit a positive (agreeing)response. The generallynegativeresultsof the a simple association reviewedresearchwould appear to contradict model prediction.However, the associativeprocess may be well or suited for simple messages in which one concept, argument, product is presentedseveraltimesin connectionwith the humor.

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

419

Messagesused in the researchreviewedwere generallycomplex, containing several arguments each presented onlyonce. Distraction effects The inclusionof humorinto a messagemay elicitlaughter which could proveto be a distraction leadingto a greater impactof message arguments(Baron et al., 1973). The failure to find increased persuasion due to humor would thus appear to decrease the likelihood that this process is an adequate description of humor's effects on persuasion. Again,more adequate tests of thishypothesis might prove to be enlightening. Accordingto the counterargument-reduction explanation of distractioneffects,for distraction to result in greater persuasion, at least two conditions mustexist: (a) recipients mustbe motivated to counterargue initially,and (b) recipientsmust be capable of counterargument. The first condition necessitates thatthe audience be opposed to the messagepositionand thatthe issue be to them.Initialopinionsof Ss werenot alwaysopposed to significant the messagepositionin the research described.Also sincethetopics were generalratherthan specificto the problemsof the audience, low involvementis likely. Finally, whether Ss could produce is unknown, since no information counterarguments is provided theirknowledge of the topics.Futurestudies, concerning the testing counterargument-reduction distraction hypothesis of humor'seffects on persuasion, might manipulate Ss' ability to counterargue via either equippingor not equipping themwithopposingmaterial priorto the of the message. receipt CONCLUSIONS Since the majority of the research reviewed has yielded non-significant results,one is faced with the question of whether humor in facthas no significant effects on persuasion, or whether the of the effect conditionsor mediumfor the emergence appropriate have not been created. Prior research is characterizedby poor methodology.More variabilityin the experimentalpopulations sampled seems imperative. Less complextopics,more conduciveto humoroustreatment, shouldbe delivered in less solemnatmospheres. Variouslevelsof attention shouldbe required in subsequent research. workshould systematically Thus, future experimental varysome of these factors. In general,the researchreviewedwas empiricalin orientation and conspicuously atheoretical.Investigators would

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

420

SOCIOMETRY

benefitfromworkingwithinsome theoretical or at a framework, froma less simplistic minimum approach to the problem. Finally, sufficient care should be devoted to increasing the power of the statistical analysesused, sincethe priorresearch findings suggest that a humor-persuasion effectis not likelyto be robust. Even a small formessages effect could be significant to the multitude. delivered
REFERENCES Baron, R., P. Baron and N. Miller 1973 "The relation between distraction and persuasion." Psychological Bulletin. 80:310-323. Berlo, D. K. and H. Kumata 1956 "The investigator: The impact of a satiricalradio drama." Journalism Quarterly 33:287-298. Brandes,P. D. 1970 "The persuasivenessof varyingtopics of humor." Paper presentedat Speech CommunicationAssociation,New Orleans. Festinger, L. and N. Maccoby 1964 "On resistanceto persuasive communications." Journalof Abnormal and Social Psychology68:359-366. Gibb, J. D. 1964 "An experimental comparison of the humorous lecture and the nonhumorous lecture in informativespeaking." Unpublished M.A. of Utah. thesis,University Gruner, C. R. 1965 "An experimentalstudy of satireas persuasion." Speech Monographs 32:149-154. 1966 "A further experimental study of satire as persuasion." Speech Monographs33:184-185. 1967a "Editorial satire as persuasion: An experiment."JournalismQuarterly 44:727-730. 1967b "Effect of humor on speaker ethos and audience information gain." Journalof Communication 17:228-233. 1970 "The effect of humor in dull and interesting informative speeches." CentralStates Speech Journal21:160-166. 1971a "An experimentalstudy of ad hominemeditorial satire: Art Hoppe vs. Martha Mitchell." Paper presented at Speech Communication Association Convention,San Francisco. 1971b "Ad hominem satire as a persuader: An experiment." Journalism 48:128-131. Quarterly 1972 "Art Hoppe vs. capital punishment:An experiment."Paper presented at SouthernSpeech CommunicationAssociation,San Antonio. Gutman,J. and R. F. Priest 1969 "When is aggression funny?" Journal of Personality and Social Psychology12:60-65. Hovland, C. I., I. L. Janisand H. H. Kelley and Persuasion.New Haven, Conn.: Yale. 1953 Communication

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

EFFECTS OF HUMOR ON PERSUASION

421

Janis,I. L., D. Kaye and P. Kirschner 1965 "Facilitating effects of 'eating-whilereading' on responsivenessto persuasivecommunication."Journalof Social Psychology1:181-186. Kennedy,A. J. 1970 "An experimentalstudy of the effectof humorous message content upon ethos and persuasiveness." Paper presented at Speech Association,New Orleans. Communication 1972a "An experimentalstudy of the effect of humorous message content University of upon ethos and persuasiveness." Doctoral dissertation, Michigan. 1972b "The effect of humor upon source credibility." Paper presented at Speech CommunicationAssociation,Chicago. Kilpela, D. E. 1961 "An experimental study of the effects of humor on persuasion." UnpublishedMaster'sthesis,WayneState University. Lampton,W. E. 1971 "The effect of humor in a persuasive sermon." Paper presented at Speech CommunicationAssociation,San Francisco. Langevin,R. and H. I. Day 1972 "Physiologicalcorrelatesof humor." Pp. 129-142 in J. H. Goldsteinand P. E. McGhee (eds.), The Psychologyof Humor. New York: Academic Press. Lull, P. E. of humorin persuasivespeech." Speech monographs 1940 "The effectiveness 7:26-40. Markiewicz,D. 1972a "Can humor increasepersuasion,or is it all a joke?" Paper presentedat Speech CommunicationAssociation,Chicago, December. 1972b "The effects of humor on persuasion." Unpublished doctoral Ohio State University. dissertation, 1973 'Persuasion as a function of humorous vs. serious messages or Northern Illinois University. contexts." Unpublishedmanuscript, McGhee, P. E. 1971 "Development of the humor response: A review of the literature." PsychologicalBulletin76:328-348. McGown, M. A. 1967 "An experimentalstudy of the persuasiveimpact of a satiriceditorial and that of a comparable directeditorial." Unpublishedmastersthesis, of Nebraska. University Orne,M. T. 1962 "On the social psychology of the psychological experiment: With and theirimplications." particularreferenceto demand characteristics 17:776-783. AmericanPsychologist D. Papageorgis, and persuasion." PsychologicalBulletin70:271-282. 1968 "Warning Pokorny,G. F. 1965 "An experimentalstudy of the impact of satiricmaterialincludedin an argumentative speech." Unpublished masters thesis, University of Nebraska.

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

422

SOCIOMETRY

Pokorny,G. F. and C. R. Gruner 1969 "An experimentalstudy of the effectof satire used as support in a persuasivespeech." Western Speech 33:204-211. Rokeach, M. 1973 The Natureof Human Values. New York: The Free Press. Rosenberg,M. J. evaluation apprehensionfrom 1965 "When dissonance fails: On eliminating attitude measurement."Journal of Personalityand Social Psychology 1:28-42. Rosnow, R. L. effect in persuasive communication?" 1965 "A delay-of-reinforcement Journalof Social Psychology 67:39-43. 1966 "'Conditioning' the direction of opinion change in persuasive of Social Psychology69:291-303. communication." Journal Rosnow, R. L. and R. E. Lana 1965 "Complementary and competing-order effects in opinion change." Journalof Social Psychology66:201-207 Rosnow, R. L. and G. Russell 1963 "Spread of effect of reinforcement in persuasive communication." PsychologicalReports 12:731-735. Smith,E. E. 1961 "Methods for changing consumer attitudes: A report of three experiments." Project Report QuartermasterFood and Container Instituteforthe ArmedForces (PRA Report 61-2). Snyder,A. F. 1971 "The effectsof demand characteristics upon conformity in judgments of humor." Dissertation Abstracts International 31:7581. Taylor,P. M. of humor in informative 1964 "The effectiveness speeches." CentralStates Speech Journal13-15:295-296. 1972 "The relationshipbetween humor and retention." Paper presentedat Speech CommunicationAssociation,Chicago. Windes,R. R., Jr. 1961 "A study of effective and ineffective presidentialcampaignspeaking." Speech Monographs28:39-49. Youngman,R. C. of the effectof germanehumor versus 1966 "An experimentalinvestigation humor in an informative communication." Unpublished non-germane masters thesis,Ohio University. Zeman,J. V. 1967 "An experimentalstudy of the persuasiveeffectsof satire in a speech presented to a high school audience." Unpublished masters thesis, of Nebraska. University

This content downloaded from 95.76.197.1 on Mon, 24 Feb 2014 06:41:23 AM All use subject to JSTOR Terms and Conditions

You might also like