hackett(1985)_hierarchy of access

Upload: ojovelador

Post on 03-Apr-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    1/12

    Bv Robert A. Hackett

    A Hierarchy of Access: AspectsOf Source Bias inCanadian TV NewsGovernment spokespersons,ranking politicians andleaders of organized groupshave better access to TV newsthan other people do.

    From gatekeeping studies to organiza-tional research, the academic literaturesuggests that news does not merely reflectthe pattern of social and political events.Rather, as an artifact routinely producedwithin specialized institutions, news ischaracterized by its own systematic orien-tations.' Such orientations may be regard-ed as "biases", so long as the concept ofbias is neither reduced to that of deliberatepartisanship, nor construed to imply thepossibility of value-free news.This article reports the results ofa con-tent analysis of television news withrespect to one type of potential bias"source bias,'" which in Stephen Kline'swords:

    . . . reflects the news agency's orientation tothe use of sources outside itself within thenewscast. The selection of who is to speakfor or about an event, how and where theyare filmed, and what they are asked to andallowed to say is an important element inshaping the range and impact of opinionthat gets expressed publicly.^

    The study was designed with two typesof source bias in mind. The first is sug-gested by interest groups (and sometimespolitical parties) who feel that coverage isdistorted, or unfavorable compared tothat of other groups with competing goals.Charges of source bias (e.g., that themedia ignore the group's views, or focuson unrepresentative members) are fre-quently part of interest group grievancesagainst the media. In recent years, laborand especially business have been two ofthe more vocal critics of news bias.' Inresponse to such criticisms, journalistsoften argue that the dissatisfaction ofcom peting interest g roups sho ws that theyare "playing it down the middle" ratherthan favoring any particular group. Bycomparing the access accorded to laborand business in the new s, the present studyaddresses the question of intergroup sourcebias.A second type of source bias is sug-gested by some academic students of thenews media. They argue that the mediarely unduly on a handful of senior electedpoliticians, and on powerful bureaucraticinstitutions and their official spokesper-sons, for information and even for a basicsense of what constitutes social reality,what counts as a recognizable and news-

    ' See. for exatnple. David L Altheide. Creating Reality: Howr News Distorts Events (Beverly Hills. Sa fe . 1476); MarkFishmati. Manufacturing the New s i\yMit\. University ot TexasPress. 19801. and Gaye Tuehtnatl. Maktng News A Study in iheConurutiion ofReality (Ne w York The Free Press. 1978).

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    2/12

    A Hierarchy of Access 257worthy event.* This orientation handicapsless formally organ ized g roup s, such as thewomen's movement, and people who donot occupy positions of institutional pow-er. The findings reported here address thisquestion also.

    MethodsTw o aspects of source bias were consid-ered: first, the types of people who appearin speech clips and the context in whichthey are filmed; and second, the level ofaccess accorded to different "actors" inlabor and business news.The national evening newscasts of the

    two major English-language Ca nadian net-works (the State-owned Canadian Broad-casting Corporation, and the commercialCTV network) were videorecorded dailyduring September and Octob er, 1980 andcontent analyzed. While the study is ex-ploratory and caution must be exercised ingeneralizing, a two-month period wasconsidered sufficient to reveal regularitiesin news reporting. The risk of selecting anatypical time period was minimized byavoiding the media's summer "silly sea-son," and by reviewing monthly workstoppage levels for the previous decade toavoid those months with unusually low orhigh activity.

    A reliability test on key variables (suchas the status and mode of each speaker)found .94 intercoder agreement. The basicunit of analysis was the discrete news item.The reported results are aggregated for thetwo networks, since differences betweenthem were generally not significant.The study was divided into two phases.In the first phase, based on the entire 1,835news items, researchers coded the identityand status of each non-media person

    ' See. for example. Fishman, op at., and Harvey Molotchand Marilyn t-ester. "News as Purposive Behavior. OiMhe Stra-tegic Use of Routine Events. Accidents, and Scandals." Ameri-can Suciological Rniew. 39 101-12(February 19741' But see the su ue sti on of Glasgow University Media Group.More Bad News (London: Routledgc A Kegan Paul. 1980). p

    whose speech was conveyed audibly andintelligibly during a news report. Multiplespeech clips from the same person in thesame item resulted in only a single coding.Status (as identified verbally by journal-ists, by supercaptions, or from the contextof the news story) was coded to combinetwo criteria: the speaker's function withinthe news story (e.g., spokesperson , expert,witness), and his or her social or politicalposition (e.g., government official, opposi-tion politician, employer, labor official,rank-and-file worker).

    In addition, the situation or mode inwhich speakers' statements were obtainedand/or presented was also coded. Thesemodes include formal public speeches,news conferences, formally arranged inter-views (in a studio/office setting, or at thesite o fa news story), unrehearsed "ad hoc"interviews, and overheard conversationwhich the speaker never intended formedia dissemination. The above men-tioned modes con stitute a descending hier-archy of access in two senses. First,although directly relevant published evi-dence is lacking,' it seems reasonable toassume that different modes convey to theviewer different degrees of authoritative-ness. For example, a formal speech or aninterview behind a desk may carry greaterweight than does an overheard privateconversation: and office interviews per-haps convey greater rationality than dooutdoor on-site ones (which, however,may carry the authenticity of directinvolvement in the news event). Second,the "higher" modes of access involve themost advance preparation by the speaker,and the least intervention by journa lists. Aspeech or news conference provides greateropportunity for a source purposefully toset the agenda of a news story than doesan interview, for example.

    The second phase of the study focusedon the 488 items which concerned theeconomy, industrial relations, or labor or

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    3/12

    2 5 8 J O U R N A L I S M Q U A R T E R L YT A B L E I

    Types of Speakers on Canadian Television News, in PercentType of Speaker Percentage of all Speech Clips"

    Canadian federal government polit icians 12.1%Canadian provincial government polit icians 14.8Canadian opposition polit icians (federal orprovincial) 8.2Canadian municipal polit icians 0.5Canadian government spokesperson (other than elected polit ician) 3.2Koreign government spokesperson or politician 8.4Foreign opposition polit ician 4.3

    T O T A L , all polit icians and government spokesper sons (51.S)"Labor spokesperson (Canadian and foreign) 5.9Business and employer spokesperson (Canadian and foreign) 3.9Spokespersons for other institutions and organized interest

    groups (Canadian and foreign) 4.3Ad hoc or non-instituiionali7ed spokespersons 2.2

    T O T A L , all non-governm ental spokespersons (16.4)Vox pops'- (ordinary people, "man in the street", including

    business and labor) 13.9Witnesses 3.0Central figures of a ncub story 3.7Other participants in a news event 6.2Experts 4.8Other (unclassifiable) 0.7

    T O T A L , all speec h clips (lOO.I)''^T w o or more speech clips by ihe same speaker in Ihe same news item were counted asone clip."Grand totals may differ from the sum of individual percentages due to rounding errors.^A sequence of rapid "talking head" appearances by Vox Pops was counted asone clip.groups or institutions) was recorded for finally, mention of an actor only in theeach news item. Multiple mentions of the context of another actor's statement,same actor or same type of actor within anitem were recorded as a single appearance. ResultsThe"highest"typeof access was coded for First Phase Findings; Speakers andeach actor-type, ranging from a speech or their Context: As Table I shows, 35. l%of

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    4/12

    A Hierarchy of Access 259is partly an artefact of the monitoringperiod. A major political story at the timewas a Constitutional Conference involvingthe federal and the ten provincial govern-ments. Journalists sought to balance theircoverage between the federal and provin-cial viewpoints, thereby ignoring the op po -sition parties which could not participatein the Conference.

    When all politicians and non-electedgovernment spokespersons are combined,the total climbs to 51.5% of speech clips(and 39% for peo ple representing govern -ments rather than opposition parties).Spokespersons for non-governmental in-stitutions or interest grou ps (Table I) alsoenjoy fairly frequent access to the news14.1% of ail speech clips. Note the low fre-quency (2,2%) of ad hoc spokespersonsrepresenting no/t-institutional clients orcolleagues. Examples during the monitor-ing period included lawyers involved inparticular court cases, and Olympic ath-letes delegated by their colleag ues to makea public statement.

    Eyewitnesses to an accident, or a centralfigure around whom the story revolvesprovide an element of human interest tothe news. Their direct involvement alsoserves to reinforce the journalist's accountof an event. Ordinary people or celebritieswho participated directly in a news eventconstituted 12.8^^ of speakers.Another avenue of access for ordinarypeople is the "vox pop" interview with a"typical" memb er of ihe pub lic, or a rank-and'file member of some broad groupsuch as a trade union. Often the vox pop'srole is to react to a news event affectinghim or her. By virtue of their typicality(rather than their elected or official sta-tus), vox pops have a certain representa-tive legitimacy and add a populist touch tothe news.

    Another route to becoming a newssource is possession of specialized knowl-edge about some newsworthy e\ent or

    with labor, employer and/or business affil-iations suggests that bias in favor of "offi-cial" sources is at least as important asbias towards any particular interest gr oup .Admittedly, labor speakers outnumberedtheir business counterparts 140 to 91,which appears to support business's com-plaints. But this finding must be qualifiedin several respects. First, labor speakerswere highly concentrated in strike stories,while business speakers were more evenlydispersed throughout various story topics.Second, internal dissension characterizedmore labor than business statements.Finally, 9 of the 71 experts interviewedhad Canadian or foreign business affilia-tions; none were trade union staff mem-bers.'

    More importantly, designated spokes-persons for particular unions, companiesor federations (such as the CanadianLabour Congress) accounted for 62.7% ofbusiness speech clips, and 57.8% of labor's(62.8%, if one includes shop stewards andstrike captains, who are usually rank-and-file workers with formal responsibilitieswithin the union). Independent business-men not acting as an expert or spokesper-son accounted for just 27.5% of businessspeech clips, while rank-and-file workersaccou nted for 37.1% of labor's (42.1%, ifone includes shop stewards and strike cap-tains as representatives of the grassrootsrather than the union hierarchy). Thereare no universally acceptable criteria toevaluate whether or not rank-and-fileworkers had sufficient opportunity tospeak for themselves. It should be noted,however, that their total ts less than that ofofficial union spokespersons, even thoughtw o of the several major labor news storiesduring the monitoring period concernedwalkouts which were not initially autho-rized by the union executive.The relatively privileged access accordedto politicians and to otficial spokespersonsis further indicated when we consider the

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    5/12

    2 6 0 J O U R N A L I S M Q U A R T E R L YTABLE 2

    Types of Speakers By Context of Their Appearance, and By Average Access ScoreType of Speaker Context of Appearance*

    AverageAccessFS NC OI OSI AH l OT O Score'' N

    Op pos ition politicians 1.77 1.29 .59 .63 .86 .39 .43 4.77 171Gov ernm ent politicians 1.49 1.54 .60 .72 .65 .43 .45 4.66 543Ad hoc spok espersons 1.06 .41 .86 1.13 3.93 4.23 31Central figure 62 1.66 .63 .99 .63 3.22 4.20 53U b o r spok espersons 42 1.96 1.22 1.06 .38 .63 4.13 77Instittjtional spokespersons(other than labor & business) . . . .61 .52 2.56 1.02 81 4.03 60Experts 06 .20 4.39 .73 1.55 3.92 64Business spok esper sons 28 43 2.38 1.29 .50 1.12 .41 3.62 59Participants in event (exc.central figure, witn ess) 94 14 .22 1.16 .65 4.37 4.02 3.53 91Business vox pops 2.94 1.52 1.32 3.36 25Labor rank & file reps (strikecaptains, shop stewards) 58 1.67 4.73 3.49 3.17 7Eyew itness 10 .33 2.28 1.44 2.38 3.08 41Vox pops (exc. labor.business) 23 .77 1.68 5.82 1.90 .20 3.00 122Labor vox pops (rank & filework ers) .73 1.78 5.04 3.18 2.78 46Overall Average (or Tota l) 24 46 15.90 149 6 34 17 3 38 3.02 4 10 4.15 1390^X^=782.I46. d.f.=78, p

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    6/12

    .4 Hierarchy of Access 261cates the newsworthiness of politicians'statements as "events" in themselves. Italso reflects the use of footage from theHouse of Commons' question period as alively means of covering Parliamentarypolitics.

    Politicians appeared less frequently inface-to-face interviews. When they did so,such dialogues tended to be "on site"rather than at a studio or office location.One can speculate that given their gener-ally high degree of access and their hecticschedules, politicians themselves choosenot to participate in extended formalinterviews. Rather, in an arrangement ofmutual convenience with journalists, theygrant brief "on site" corridor interviews(known as "scrums" in Ottawa press jar-gon) to discuss recent proceedings, whileformal speeches and news conferences areused for major polic>' announcements.

    Not surprisingly, most vox pops andespecially eyewitnesses were interviewedon site, and experts spoke in office or stu-dio settings.Labor, business, and other institutionalspokespersons were under-represented informal speeches. Non-labor spokesper-sons typically appeared in office or studiointerviews. Interestingly, both businessspokespersons and vox pop businessmenappeared in more studio/of^fice interviewsthan their labor counterparts. Labor offi-cials were more likely to give news confer-ences, while rank-and-file workers ap-

    peared in more on-site interviews, over-heard conversations and other situations.The most plausible explanation for thispattern is that a great many of the laborspeakers, both leaders and rank-and-file,appeared in connection with a strike byfederal government clerks. The unionleaders needed rapid access to the media inthe contest for public op inio n, partieularlyimportant in public sector dispu tes; but asa general principle, they wanted to avoidgiving exclusive personal interviews which

    ers' press conferences, television coverageof the strike focused on the clerks' picketlines and rallies, resulting in (on site)interviews with rank-and-file strikers,chanted slogans (other mo de), and debatesamong clerks or between clerks andmembers of the public (overheard talk).

    By contrast, no single news event in-volved corporate officials or businessmenin the same way. Their news conferenceswere not urgently relevant to breaking sto-ries; their activities were less photogenicthan picket lines. With business state-ments considered relevant to a more dif-fuse and less pressing set of events andissues, journalists and businessmen couldafford the "luxury" of personal interviewsin corporate offices.

    The average access scores for each typeof speaker, reported in Table 2, must beinterpreted with considerable caution,since they result from treating ordinal dataas if they were interval. Nevertheless, theydo provide a rough confirmation of a hier-archy of access which favors politiciansand official spokespersons. Oppositionand government sources enjoyed on aver-age the highest forms of access. At the bot-tom of the totem pole were all vox popgroups (especially rank-and-file workers),strike captains and ordinary people whohappened to witness a news event.Second Phase Findings: Actors in Eco-nomic and Industrial News: While thesecond phase data are based on a smallernumber of items than are the first phasedata, they include all actors which arementioned in news items, whether or notthey speak during the item. Therefore, thesecond phase findings enable us further tospecify the level of access accorded to dif-ferent groups.Table 3 shows the distribution of thehighest form of access obtained by each offive broad groups in each news item aboutthe economy, tabor, business or industrialrelations. Once again, government and

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    7/12

    2 6 2 J O U R N A L I S M Q U A R T E R L YlABLE 3

    Groups in Economic/Industrial Items in TV News,By "Highest" Level of Access in Each ItemGroup

    State agen-Level of Acce ss Politicians, cie s(e xc l. Allparties, gov't Labor politicians) Business othersN of items in which groupappears (out of 48 total) 218 227 227 254 284Mentioned by non-joumalistonly 6.4% 2.6% 4.8% 7.5% 10.9%M entione d by journa list 7.8 3.1 9.7 5.] 9.2Participant, but no quote

    or speech clip 31.7 46.3 45.8 54.3 51.8Sp eech clip overheard talk 1.8 0.4 0.4 0.7Attr ibuted qu ote 17.9 13.2 24.7 12.2 10.6Speech clip: formal speechor interview 36.2 33.0 14.5 20.5 16.9

    To tal ("r 100.0 100.0 99 .9 100.0 100.1Aver age acce ss score per item* 4.24 4.17 3.74 3.66 3.42X'=IO4.I2. d.f.=20, p

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    8/12

    A Hierarchy of Access 263TABLE 4

    Relative Frequency and Degree of Access of SubgroupsIn Econ om ic Industrial New s Items, in Percent

    Group

    Politicians, Parties, GovernmetttFederal governmen t. C abinetProvincial GovernmentFederal Conservative party (the official opposition),.,Federal New Democratic Party (opposition party)Other politicians or parties

    State Corporations, AgenciesState corporation, public sector in role as etnployer..,State as administrator or provider of servicesPublic sector business, other than in employer role . . .State enquiry, mediator (e.g. labor relations board)..,Police, armed forces, judiciary

    Labor GroupsUnion m embership; group s of rank-and-file unionists .Specific trade unions or their leadersWorkers in general; occupational groups, notnecessarily unionizedOrganized labor in getieral; trade union federationsand their leadershipStrike captains, shop stewards

    Business GroupsPrivate sector business; general categories such as"corporations", "investors", "executives", etcSpecific company and its spokespersons, other than inrole as employer or manager of laborSpecific company and spokespersons, as employerSmall businessmen , farmers, professionalsCorporate associations, pressure groups

    All Other Groups"Canada"; the public; consumers; averagetaxpayer, etcForeign economic and political institutionsExperts, research institution s, etc ^The news media; journalistsThe unemployed; laid-off workersOrganized pressure groups, other than thoserepresenting groups listed elsewhere (i.e. business,

    Percent of GroupPercent of Items App earances whichwith each type are Interviews orof group Quotes

    (N=488)3 6 . 32 6 . 26.83.59.6

    1 7 . 61 3 . 7I I . 51 0 . 29.4

    3 2 . 82 2 . 3

    1 8 . 9

    5.70.8

    384355829

    3736214015

    1972145775

    37.1 101 7 . 88.86.83.1

    3 8 . 91 5 . 69.69.43.3

    47283693

    912682013

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    9/12

    264 J O U R N A L I S M Q U A R T E R L Ysignificant at .01, with the exception ofbusiness versus "other g roups" But thewithin-group differences are at least equal-ly revealing. Table 4 shows the frequencywith which various sub-groups appear ineconomic or industrial news items, alongwith the percentage of such items contain-ing an interview/speech clip and/or anattributed statement from each sub-group.This latter statistic highlights the distinc-tion between sub-groups who appear inthe news as speaking "subjects", and thosewho appear as "objects" whose actions aredeemed newsworthy (sometimes becausethey are perceived as threatening), butwhose statements, rationales, or view-points are not.As the table shows, federal and provin-cial governments and their spokespersonsappear both frequently and with a highdegree of access, indicating their status asroutine sources of information and view-points. The prime minister and his severaldozen colleagues in the Cabinet appearedin 97 of the 177 items in which the federalgovernment was mentioned. The far morenumerous backbench members of Parlia-ment, and the entire extra-ParliamentaryLiberal Party ap para tus, appeared in justsix items. It seems that within each group,a small number of high-status individualsrecurringly appear as its representatives,thus further sharpening the apex of thepyramid of access.For reasons suggested previously, thefederal opposition parties are well downthe list, as Table 4 shows. Note, however,the similarity of the federal government'stotal with the combined provincial andopposition total.Other political forces, such as provincialopposition parties, municipal politiciansand extraparliamentary parties, are rela-tively absent, due no doubt to the Parlia-mentary and national focus of networktelevision journalism. They are mentionedabout as often as are federal opposition

    the leaders of specific unions are wellabove average in this respect. These resultssuggest that journalists focus on ordinaryunionized workers by virtue of theirnewsworthy actions, whereas leaders aresought for their explanations of the rea-sons and future course of such actions.Reporters in effect assume that unionleaders can and do speak for theirmembershipan assumption which isespecially problematic in the case ofserious divisions within the union, and/orwildcat walkouts.Private-sector business groups appearin more economic/industrial news itemsthan does labor, as Table 3 shows. How-ever, only 9.9% of all business appear-ances were in stories about labor-manage-ment disputes.'* Not surprisingly then,business usually appeared in roles otherthan that of a protagonist in labor rela-tions (employer or manager). Even whenpublic sector managers are included, busi-ness actors as employers are outnumberedby trade unionists.The most frequently mentioned sub-group is the public, similarly broad cate-gories such as consum ers, Canadians, tax-payers and commuters, or more specificgroups of o rdinary citizens affected by anews event, such as an a irport strike or thegovernment's budget. However, membersof these groups rarely speak or get quoted.Moreover, in 18% of their appearances(compared to an average of 10% for allsub-groups), they are mentioned only by anon-journalistic speaker, and ignored inthe news script itself. (This result derivesfrom the crosstabulation of sub-groups bytypes of appearance, not reported in full inTable 4.) Transcripts of the economic/in-dustrial news items suggest that "the pub-lic," "the people of this country," and sim-ilar labels are often invoked by speakersseeking democratic legitimacy to supporttheir case. In the news, the public isinvoked or acted upon, rather than acts orspeaksunless one accepts that interest

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    10/12

    A Hierarchy of Acce.ss 265or the poor, which are generally unrepre-sented by organized interest groups, withthat of organized interest groups. The lat-ter appear with approximately equal fre-quency but are far more likely to providequotes or interview/speech clips. Thisfinding suggests the importance of access-ible spokespersons and institutionalizedroutines for generating information ascriteria of access to the news.

    DiscussionThe labor-business comparison doesnot reveal obvious source bias in favor ofone of the two groups. Business outnum-bered labor as an actor in second phasenews items, but labor speakers appearedmore frequently overall, especially inindustrial dispute stories. One possibleexplatiation for this pattern is a tendencyof corporate leaders to "hide" behind asingle public relations official, whereasduring a strike, the media have access tomany trade unionists, who have the addedadvantage of "human interest" newsw orthi-ness." Quite another interpretation isoffered by Hartmann:

    Because the legitimacy of union activityis suspect, the unions are more frequentlycalled upon by the media to account fortheir actions or feel obliged to make state-ments in their own justification. It wouldseem that emp loyers, assured in their legit-imacy, seldom feel this need and that themedia accept this situation as normal.'^Any source bias between interest groupsor between major Parliamentary politicalparties seems to be overshadowed by thehigh degree of access for governmentspokespersons, and for ranking politiciansfrom the federal opposition as well as thefederal and provincial governments. Highaccess is indicated by the frequency andthe mode of their appearances as bothspeakers and actors in the news. These

    i< A H. Raskin. "Double Standard or Do uble-Talk'^ in Aro-noff. op. n( ..p . 2 il .

    results suggest in part the importance ofgovernment and politics as a news topic,but one may also infer that politicians andgovernment officials satisfy many of theconsiderations that journalists apply whenselecting sources: accessibility, involve-ment in predictable conflict, ability tosupply useful knowledge, power over thecourse of events and assumed representa-tive status.'}

    Similarly, the leaders or designatedspokespersons of organized groups andinstitutions enjoy relatively privilegedacces s. This pattern can likewise be attrib-uted to the factors of sources' availabilityand preferences in dealing with the media,and to journalists' efficiency-oriented judg-ments of sources' suitability.Ordinary people are certainly not absentfrom the news, but they are dispropor-tionately merely mentioned rather thanquoted or interviewed. When they doappear as speakers, it is disproportion-ately in a lower status context.These patterns may have importantunintended consequences for the flow andnature of public information, since asTuchman argues:Whom one asks for information influ-ences what information one receives.. . .They (journalists] contacted the powerful,the politician with the rt:sources to accom-plish his or her ambitions, not the merelydissident or dissatisfied. That people withpower serve as sources bears consequencesf'or the information newsworkers un-cover. ... The choices of sources and thesearch for "facts" mutually determine eachother. Together, .. .they flesh out the newsframe.'*

    Though the data reported here are notconclusive (since one would need to exam -ine what is said on the news as well as whosays it), they are consistent with the claimof Hall and colleagu es that the "accreditedrepresentatives of major social institu-tions" become the "primary definers" of

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    11/12

    Reporting of Presidential Puh lic Opinion Polls 277(MOR) of Detroit to do its polling, andoften Frederick P. Currier, president ofthe organization, wrote the story. PhilipMeyer, who worked for Knight Newspa-pers during some of the years examined inthis study and later wrote a book on pub-lic opinion polling aimed primarily forjournalists and journalism students, wroteseveral poll stories for the Free Press Hefrequently sought the help of AngusCampbell, of the Survey Research Centerat the University of Michigan, and citedhim in hii stories. During 1980 and 1984.the .Vfit \ featured copyrighted polit icalpoll stories by Michael Traugott of theinstitute for Social Research at the Uni-versity of Michigan. The sharp increase inthe reporting of response rate during 1980and 1984 can be largely attributed toTraugott's stories.

    Even though this study suggests that thereporting of public opinion polls duringpresidential election years has improved,the reporting of a few variables can stillstand improvement. Even in 1984 only aquarter of the poll stories reported ques-tion wording. Even many of Traugott'sstories did not include question wording.

    '* Perso ful inlerMtw with Michael Traugott *\ the CentralMichigan Chapter of Ihe Socieiv ol Proleskional JournalislkSigma Delta Chi. U ns in g. Mich So t 14.19114

    According to Traugott, editors believequestion wording takes up valuablespace.'" From 1968 to 1980. most storiesdid not report sampling error. While thereporting of sampling error does not takeup much space, perhaps the editors feltthat the inclusion of sampling error mightconfuse readers. The reporting of responserate was abysmally bad until 1984, v^henalmost a quarter of the polls includedresponse rate. The reporting of responserate is not required by the AAPORguidelines.

    Future research should investigate wheth-er metropolitan dailv newspapers doindeed regard in-house polling as prcsti-gous. Research should also e.xamine howsmall- and moderate-si/e newspapers re-port public opinion poll sto ne s Longitud-inal research should further investigatewhether the indications of change found inthis study apply to other m etropolitan dai-lies. Finally, research should mo\e av^a\from examining the poll as the unit ofanalysis and examine audience members"perceptions of poll stories. In particular,e.xperimental research should investigatewhether the inclusion of needed methodo-logical information in poll stories leads toincreased understanding and credibilityamong readers.

    A HIERARCHY OF ACCESS

  • 7/28/2019 Hackett(1985)_Hierarchy of Access

    12/12