A content analysis of alcohol portrayal and drinking in British television soap operas

HEALTH EDUCATION RESEARCHTheory & Practice

Vol.12 no.4 1997Pages 519-529

A content analysis of alcohol portrayal and drinking inBritish television soap operas

Adrian Furnham, Henriette Ingle, Barrie Gunter and Alastair McClelland

Abstract Introduction

This study examined the portrayal of alcoholconsumption in a sample of six soap operas onBritish television. In all, 86% contained visualor verbal references to alcoholic beverages withan average of one reference in every 3.5 min ofprogramming. Over all of the 25 programmesmonitored, more alcohol was consumed thansoft drinks, at a ratio of 2.1:1, and there werehardly any references to the hazards of alcoholconsumption. In contrast to previous similarstudies, alcohol was found to be most promin-ently featured in Emmerdale Farm and hardlyever appeared in Home and Away. Also, four ofthe major soaps regularly showed specificbrands of alcohol, in particular Eastenders.Finally, despite the observations in previousstudies by others that the presentation of alcoholon television is decreasing, in comparison tothree studies done in the last 10 years, alcoholportrayal appeared to be, if anything, on theincrease, although problems associated withthese comparisons render the results in need ofreplication. The quality of data yielded in manyof these analyses, along with poor base-ratedata, make it difficult to establish the evidenceof reliable trends, though preliminary analyseswere done comparing two earlier studies withdata obtained in this one. However, recom-mendations for further studies have been made.

Department of Psychology, University College London,26 Bedford Way, London WC1, UK.

With increasing concern over the prevention ofalcohol abuse, attention has turned to the questionof whether references to, or the portrayal of, alcoholconsumption on television, either in programmes oradvertising, has an effect on consumption ofalcohol. Both advertising and television programshave been criticized for providing an image ofalcohol that is generally inconsistent with thereality of the problem it causes in society (Atkinet al., 1984; Wallack et al., 1987). There is aconsiderable and increasingly sophisticated literat-ure on media effects which has focused mostspecifically on the impact of sexual and violentmaterial (Bryant and Zillmann, 1994). The effectsare considered to be cognitive, affective and beha-vioural, but there is considerable debate concerningmethodological problems surrounding these studies(Hanneman and McEwan, 1976). Brown andWalsh-Childers (1994) noted that two questionshave been addressed in the alcohol-related literat-ure: does alcohol advertising encourage non-drinkers (especially adolescents) to drink? and doesit encourage drinkers to drink more or in moredangerous circumstances as before? They believethe available research shows advertising alcoholdoes indeed lead to both of the above. However,this increasingly sophisticated methodologicalliterature has not focused very much on the relation-ship between media/television portrayals of alcoholconsumption and alcohol-related problems insociety.

Most of the research into alcohol portrayals inthe media has focused on advertising, and has paidlittle attention to the images of alcohol and drinking

© Oxford University Press 519

by guest on February 26, 2013http://her.oxfordjournals.org/

Dow

nloaded from

http://her.oxfordjournals.org/

A. Fumham et al.

behaviour seen in the media in general. Commercialadvertising of alcohol accounts for only a relativelysmall proportion of the overall alcohol images ontelevision. Furthermore, since programmes occupyconsiderably more space than advertisements theyare potentially a much greater source of reference.The Department of Health and Social Securitydocument of 1981 entitled Drinking Sensibly sug-gested that, 'It would not be surprising if thedepiction of drinking during TV and radio pro-grammes, in plays, in books and in films, werea stronger influence on people's behavior thancommercial advertising by obviously interestedparties'.

Interest in the portrayal of alcohol consumptionon television stems in a large part from concernsabout the allegedly negative effects that televisionprogrammes can have on viewers, e.g. violenceand drug abuse (Feldman, 1980). It has beensuggested that if television continually shows alco-hol consumption in a positive light, such contentmay contribute towards increased consumptionlevels. This effect may be especially likely to occurif drinking alcohol is depicted as an important partof an attractive and successful lifestyle (Wallacket al., 1985, 1987).

USA and Canadian studiesThe majority of studies on prime-time televisionhave found that alcohol use is frequently shownand that depictions of heavy drinking are common,while any presentation of alcohol-related problemsis relatively rare. For example, alcohol has beenfound to be the most frequently used 'drug' ontelevision and the most frequent beverage seen tobe consumed (Breed and De-Foe, 1978; Cafisoet al., 1982). Studies looking back at alcoholportrayals on television in the 1970s showed ageneral pattern of alcohol portrayal as being gla-morized, alcoholic intoxication as humourous anddrink being used as a coping mechanism (Hanne-man and McEwan, 1976; Atkin and Block, 1981;Breed and De Foe, 1981; Breed et al., 1984).

Studies focusing on the alcohol content of day-time serials (soap operas) found drinking levelssubstantially below those in programmes shown

during prime-time (Lowry, 1981). Wallack et al.(1985) carried out an in-depth study which sug-gested alcohol portrayals in day-time soaps pre-sented a general view of alcohol as a potentiallyproblematic substance.

Possibly the most extensive analysis of drinkingbehavior on television was a 9-year project carriedout by Wallack, Breed and De Foe, which startedin the mid 1970s. A paper by Breed and De Foe(1981) focused in detail on prime-time televisionprogrammes such as situation comedies and serialdramas. On the basis of viewer ratings, the top 15sit-coms and 15 dramas were selected from the1976-77 period. Ten episodes of each of the sit-coms and five episodes of each of the dramas wereanalysed, in total 150 hours of air time. In all,72% of the drinkers were men and the majoritywere 'settled adults' (people with family and acareer). Generally, the majority of characters seento drink were male, 'settled', 'positive' characters,usually playing the leading roles and seen at leastweekly by millions of viewers. 'Bad guys' wererarely seen to drink. Over half of the drinkingportrayed was potentially problematic: over 40%of the drinking scenes recorded involved heavydrinking and an additional 18% involved chronicdrinkers.

Breed et al. (1984) showed how levels ofdrinking portrayal have changed over the past 32years on American television and, in particular, onsoap operas. Their findings showed that alcohol-related acts in situation comedies increased from1.28 per hour in the 1950-63 period to 8.25 perhour in 1981-82 and in dramas from 4.52 to 8.95per hour. The latest published update, by Wallacket al. (1987), found alcohol to be present in60% of all prime-time television, meaning then inaudience terms, a regular viewer of dramas wouldbe likely to see more than 20 alcohol-acts perevening. This suggests a steady overall increase inalcohol portrayal in popular programmes which issimilar to that found in studies in the UK (Pendletonet al., 1991).

British studiesOver the past 10 years, a number of studies haveconcentrated on the systematic analysis of alcohol

520


Dow

nloaded from


A content analysis of alcohol portrayal

portrayal on British television, most notablyHansen (1984) and Pendleton et al. (1991). Han-sen's work looked at all prime-time viewing overa 2-week period broadcast by all four (as then)channels of British television. He found thatapproximately two-thirds of the programmes(including drama, sit-coms, films, news, document-aries and other factual programmes) contained atleast some visual or verbal references to alcohol,while in nearly one-third of the programmes,alcohol was seen to be consumed by a speakingcharacter. While fictional programmes containedover three individual scenes per hour of charactersdrinking alcohol, the average occurrence in thepopular British soap operas was considerablyhigher at almost five scenes per hour. Overall, the'drinking scenes' constituted 14% of the totalprogramme time of British soap operas, and 10and 6%, respectively, of the programme time ofAmerican-produced and Australian-produced soapoperas. However, as Hansen pointed out, this isnot particularly surprising when the majority ofthe British soaps are based around a pub.

Hansen (1984) argued that television tends tocultivate the belief that alcohol consumption isthe norm, rather than the exception in socialinteractions, and that alcohol is often a mark ofaffluent lifestyles and achievements. It also tendsto mask the fact that alcohol consumption cancontribute to accidents, violence or even death.

Two similar studies were carried out by Smithet al. (1988) and Pendleton et al. (1991), coveringtelevision output samples in 1986 and 1988,respectively. Both of these studies based theirmethods on the content analysis procedures usedin Hansen's (1984) study, thus enabling at leastsome comparison across the studies. In both studiesthey carried out a systematic analysis of 50 pro-grammes taken from the national top 100 pro-gramme chart published in Broadcast magazine.In the first study, they found that four out of fiveprogrammes contained visual or verbal referencesto alcohol and that on average there was a referenceto alcohol in every 6 min of programming. Theprogrammes were also found to show more alcoholbeing consumed than soft drinks and there were

few references to the hazards associated withalcohol consumption.

The second study found that three-quarters ofthe programmes contained visual or verbal refer-ences to alcohol and that on average there was areference to alcohol in every 6.5 min of program-ming. They also noted that alcohol was mostprominently featured in fictional programmes, withEastenders emerging as one key example. One-third of the programmes showed specific brandsof alcohol. In comparison to their earlier study,however, there appeared to be fewer alcoholimages overall.

The aim of this study was to obtain a moredetailed account of the extent and nature of alcoholportrayals in soap operas currently showing onBritish television. One problem with content ana-lysis studies has been a wide variance in the codingframes used to analyse particular programmethemes, rendering many of the findings non-com-parable (Kolbe and Burnett, 1991). For this reason,it has been difficult to assess any change in alcohol-related portrayals on television over the yearsand therefore to establish whether the programmemakers are making any effort to control the levelof such portrayal or at least to present a morerealistic image of alcohol consumption. This pre-sent study aimed to overcome this problem byusing the same coding procedure as used byHansen (1984), Pendleton et al. (1991) and Smithet al. (1988).

This study looked at the extent of non-alcoholicbeverage consumption in terms of visual and verbalreferences, and the number and duration of scenesfeaturing alcohol and non-alcohol beverages con-sumption.

The study attempted to answer the followingquestions:

(1) How much portrayal of alcohol consumption(and non-alcoholic beverage consumption) isthere in current soap operas and how is thisportrayal distributed across the different majorBritish soap operas?

(2) How much portrayal is there of the possibleeffects and consequences of alcohol consump-

521


Dow

nloaded from


A. Furnham et al.

tion—do the programmes accurately representthe problems associated with heavy drinkingand alcoholism?

(3) Who drinks where, how much, in what contextsand for what purposes?

(4) How does current soap opera portrayal ofalcohol compare to that seen in past studies(Hansen, 1984; Smith et al, 1988; Pendletonetal., 1991).

Method

Programme sampleSix soap operas were selected, all of which werein the top 20 viewing figures for 1994 (Gunteretal., 1994). They were Neighbours (BBCl), Homeand Away (ITV), Eastenders (BBCl), CoronationStreet (ITV), Emmerdale Farm (ITV) and Brook-side (C4). Inevitably, the choice of the soap operas(some, in fact, centred around pubs) must restrictgeneralization of this data. Five episodes of eachsoap opera were videotaped between December 1,1994 and January 30, 1995. In cases where therewere two episodes in 1 day, the evening showingwas always taped. All taping, therefore, was con-ducted between 17:35 and 20:30. Of the 30 soapopera episodes recorded, only 25 were analysedbecause of various anomalies concerning theremaining five, such as changes in the schedule ofa particular programme. This sample was felt tobe sufficient as soap operas are shown the wholeyear round, though it is possible that some pro-grammes taped over the festive period showedscenes where the consumption of alcohol washigher than at other times of the year. It should benoted, however, that none of the scenes analysedwere of Christmas parties.

Coding procedureThe coding procedure used was modelled on thatemployed by Hansen (1984), although it wasmodified so that non-alcoholic drinks were codedin the same way as alcoholic drinks. In other words,when people consumed non-alcoholic drinks, thisinformation was coded according to the samecriteria as when they consumed alcohol. The main

units of analysis were visual and verbal referencesto alcohol (both the beverage and drinking), andthe duration of the drinking scene. A drinkingscene was defined as any scene showing the actualor implied consumption of alcohol by a speakingcharacter; a visual reference was defined as ascene showing alcoholic drinks, their containersor promotional materials; a verbal reference wasdefined as an uninterrupted sequence of dialogueabout alcohol or its use, effects or properties. Inaddition to this, each speaking character seen toconsume alcohol was coded on the basis of smoker,sex, age, social class, occupation, drink type,location of drinking, attitudes of others and effectsof drinking.

Coding reliabilityInter-coder reliability was tested to make surethat the data were not contaminated by coderidiosyncrasies. A subsample (five programmes)was taken from the main sample of soap operasand double coded independently by one maleand one female coder. Inter-coder reliability wasassessed and was found to be high (97.5%) for theoverall episode coding and slightly lower (92%)for individual character coding. Thus coder A hada 58 versus 60 split between the middle andworking class of the characters, while for coder B,this was 66 versus 62 (see Table VI). The highestagreement was total concerned types of drinks (seeTable V) and location of the characters drinking.Differences were reconciled after discussionbetween the two coders so that total agreementwas obtained.

Results

Within this study, alcohol was referred to eithervisually or verbally in 86.7% of the programmes,while in 73.3%, the actual or implied consumptionof alcohol was shown.

Altogether, there were 96 drinking scenes, 145verbal references and 140 scenes with visual refer-ences to alcohol. Table I shows the comparison ofthe alcohol drinking references to non-alcoholic(soft drink) references. Altogether, 69% of thedrinks seen to be consumed were alcoholic.

522


Dow

nloaded from



Table I. Comparison of the number of alcohol drinking scenes and references in comparison to non-alcohol (rate per hour)

Type of drink Drinking scenes Visual references Verbal references Total references

Number of alcoholic drinkscenes and referencesRate per hourNumber of non-alcoholic drinkscenes and referencesRate per hour

96

6.444

2.9

140

9.356

3.7

145

9.6101

6.7

285

19157

10.5

There were over twice as many alcohol as non-alcohol drinking scenes and visual references toalcohol and, although the verbal references to non-alcohol were quite high, they were still far fewerthan the verbal references to alcohol. The drinkingscenes amounted to 15.2% of the total programmetime (15 hours) and occurred on average 6.4 timesper hour. The total hourly rate of references toalcohol was 19. Thus, on average, a viewer ofpopular soap operas would encounter a referenceto alcohol every 3.2 min, compared to every 6 minfor non-alcoholic drinks. It should, however, bepointed out that a rate-per-hour measure of drinkingscenes does not tell us how much any givencharacter drinks. This is clearly very important,but difficult to rate unless it is the intention of thescriptwriter to indicate a character is drunk.

Comparison with previous studiesIn general, the results from this study reveal agreater number overall of alcohol portrayals thanthat found in the studies by Hansen (1984) andSmiths al. (1988).

As can be seen in Table II, both drinking scenesand verbal references occurred far more frequentlyin this study than in any of the other three. Furthervisual references, although occurring at almost thesame rate as in both of the Pendleton et al. (1991)and Smith et al. (1988) studies, were almost twiceas frequent as the rate found by Hansen (1984).Table III is, however, more suitable for comparison,as Smith et al. (1988) and Pendleton et al. (1991)collected specific data on alcohol presentation inEastenders, using approximately the same methodsthat this study used.

As can be seen from Table III, the total references

to alcohol appears to have increased slightly overthe last 10 years, though various differencesbetween studies suggest these comparative resultsare, at best, tentative. Visual references to alcoholappear to have fallen very slightly. However, giventhe small number of programmes involved andslight differences in the sampling, one has to bevery cautious about interpreting the comparativeresults.

It is important, but difficult, to show any trendis reliable because various indices (scenes, refer-ences) must overlap to some degree.

In both Smith et al. (1988) and Pendletonet al. (1991), 150 min of broadcast time of theprogramme Eastenders were analysed, as was thecase in the present study. The actual frequenciesof occurrences (e.g. drinking scenes) were notreported by either Smith et al. or Pendleton et al.,but could be calculated by dividing the rate perhour by 60 and multiplying by 150. Using thesedata, together with current results, expected fre-quencies (under the null hypothesis of no changein rate over the years 1986, 1988 and 1995) werecalculated. An extension of the binomial test forthe comparison of Poisson rates (Snedecor andCochran, 1989) was then applied to the data. Therewas no statistically significant effect of year onthe rate of occurence of either drinking scenes orvisual references to alcohol (x2 n o t significant);however, there was a significant change in verbalreferences from 8.1 to 9.4 to 16.0 per hour overthe three studies (N = 84: X\D = 800, P < 0.02).There was also a significant change in the totalnumber of references to alcohol from 19.0 perhour in 1986, to 27.0 per hour in 1988, to 30.0per hour in 1995 (N = 191: x\2) = 6.17, P <

523


Dow

nloaded from


A. Fumham et al.

Table II. Comparison of alcohol portrayals in 'all fiction' over the last 10 years

All fiction Current study(soaps only 1995)

Smith et al.(all fiction 1988)

Pendleton et al.(all fiction 1991)

Hansen(all fiction 1984)

NumberMinutesDrinking scenes (no.)Rate per hourVisual references (no.)Rate per hourVerbal references (no.)Rate per hourTotal references (no.)Rate per hour

30900966.4

1409.3

1459.6

28519

4.5

10

5.1

15.1

5.4

9.2

5.7

14.9

27900

583.9

875.8

1208

20713.8

Table III. The prevalence of alcohol in Eastenders over the last 10 years (rate per hour)

Eastcndcrs only Current study Pendleton et al. (1991) Smith el al. (1988)

Year of broadcastNumberMinutesDrinking scenes R/HVisual references R/HVerbal references R/HTotal references R/H

1994/1995(n = 5)150

9.2141630

1988

6.417.69.4

27

1986

6.410.98.1

19

0.05). With respect to verbal references, it wouldappear that there was a major increase between1988 and 1995, whereas for total references thebiggest change was between 1986 and 1988.

Finally, it was possible to examine the prevalenceof particular alcoholic and non-alcoholic drinksover the three studies done in 1986, 1988 and 1995.There was no significant difference in references towine, spirits or non-alcoholic drinks, but there wasan increase in references to beer (2.6, 2.9, 5.2,X2<2) = 21.39, P < 0.001). There was a significantincrease in total alcohol references which can beattributed to an increase in references to beer (6.7,5.9, 9.37; x\i) 16.21, P < 0.001).

Channel of broadcast

Alcohol image frequency did not seem to differdramatically between BBC1 and ITV, with soaps onBBC1 showing approximately 12 alcohol drinkingscenes per hour, ITV 12.8 drinking scenes perhour, but Channel 4 appeared to be much lower

with only 7.4 alcohol drinking scenes per hour.This is no doubt because Brooks'ide is centred ona housing estate whereas other UK-produced soapscontain a great deal of action in pubs.

Country of origin

A large difference was seen between the British-made soap operas and the Australian ones (TableIV).

Using the same method, it was also possible tocompare soap operas set and made in Britain withthose from Australia. The analysis was based ona total broadcast time of 600 min for the Britishsoap operas and 300 min for the Australian soapoperas. With respect to drinking scenes, there wasa dramatic and highly significant difference, withBritish soaps having an average of 8.9 alcoholscenes in comparison with only 1.4 per hour forthe Australian soaps (N = 96: X2{\) = 29-30> p <0.001). References to alcohol were also much morefrequent in British soaps than Australian soaps

524


Dow

nloaded from



Table IV. Comparison

Country of origin

Australia (n = 2)Britain (n = 4)

of drinking scenes (rate per

Alcohol scenes

1.48.9

hour) between

Non-alcohol

2.83

British and Australian soap operas

scenes Total referencesto alcohol

4.826.1

Total referencesto non-alcohol

9.89.5

(26.1 versus 4.8 per hour), a difference which wasalso highly significant (N = 285: x2

(i) = 79.60, P< 0.001). There was no significant difference in thefrequency of non-alcohol scenes or in references tonon-alcohol.

Audience size and alcohol content

Based on figures collected by the Broadcasters'Audience Research Board (BARB), alcohol didtend to be more prominent in the programmes withthe biggest audiences. However, Emmerdale Farm,the soap opera found to have the highest alcoholcontent, had the second lowest viewing figures ofall the soaps chosen (13.51 million). The Emmerd-ale Farm audience remains high; besides it isshown earlier in the evening than Eastenders andCoronation Street when audiences generally arelower. However, all of the soap operas have 'high'viewing figures in comparison to other pro-grammes, and so to see the relationship betweenaudience viewing figures and alcohol content, awider range of programmes would be needed(Gunter et ai, 1991).

The type of drinks being consumedIt was also possible to examine the prevalence ofparticular alcoholic and non-alcoholic drinks visualreferences over the three studies conducted in1986, 1988 and 1995. This analysis was based on1614 min of broadcast time in the 1986 study,1463 min in the 1988 study and 900 min in thepresent study. No statistical significant differenceswere found in the frequency of references to wine,spirits or non-alcoholic drinks (all three x2 analysesnon-significant), but there was a change in thetotal references to alcohol over the three studies(6.7, 5.9 and 9.3 per hour) which proved to besignificant (N = 464: x2

(2) = 16.21, P < 0.001)and which can be attributed to the significant

increase in references to beer (2.6, 2.9 and 5.2 perhour; N = 219: x\i) = 21.39, P < 0.001). SeeTable V.

Characters seen to consume alcohol

Table VI shows the basic demographic features ofcharacters shown consuming alcohol in the soapoperas. In agreement with past studies, the majorityof drinking characters were male, outnumberingfemale drinkers by a ratio of 1.7:1. The age ofdrinkers was predominantly in the 25—44 range(48.1%), with the rest of the drinkers distributedfairly evenly among the other age groups. Incontrast to Hansen's (1984) study, a percentagewere also in the over 65 group and there were nodrinkers obviously under 18. However, quite alarge percentage (18%) of the characters were inthe 18—24 age groups. Social class of the drinkingcharacters appears to be roughly divided betweenthe working class and middle class with fractionallymore characters in the latter. Overall, from thedata collected, over 60% of the alcohol actsinvolved main characters from the soaps.

Location of drinking

Characters were generally seen to be drinking ina public drinking area (N = 65) (49.6%) or at aprivate party (N = 29) (22.1 %), but quite a largenumber (16.8%) were shown drinking at home aspart of their daily life and (N = 12) 9.2% wereshown drinking while at work. The main reasonsfor drinking were general socializing or celebration.In all, 11 characters were shown to be drinking toescape from reality, but this was incidental to themain story. Characters did not automatically acceptdrinks when they were offered them and were notalways seen to consume them. A total of 13characters were recorded as having refused drinksand as a percentage of the total number of times

525


Dow

nloaded from


A. Furnham et al.

Table V. The number of visual references to different types of drinks (rate per hour)

Type of alcoholic drink Visual references

1995-current 1988 1986

BeerWineSpiritsTotal alcoholic drinksTotal non-alcoholic drinks

5.23.42.39.36.7

2.92.725.97.1

2.62.92.66.76.7

Frequency (1995): beer (n = 62), 47.30%; wine (n = 42), 32.10%; spirits (n = 25), 19.10%; other (n = 11), 1.50%.

general population (Gunter et al., 1994). Followingearlier research, criticisms have been made of thetelevision industry for the frequency and inaccur-acy with which it portrays alcohol consumption.The current study suggests that television continuesto provide a biased portrayal of alcohol consump-tion. A comparison with several past studies indi-cated that if anything, the overall reference toalcohol, alcohol-related scenes and beer drinkingis even higher in soap operas than in past years.Indeed, extreme caution is urged before theseresults and studies of this kind are used to informpublic policy.

There are all sorts of subtle issues that contentanalysis of this kind has not addressed. Forinstance, the portrayal of alcohol as essentiallypositive or negative. More importantly is the issueof baselines. Do soap operas accurately reflectdrinking in society? If more people drink alcoholin pubs than drink soft drinks, why should it beotherwise in 'soap pubs'? Indeed, it is possible thattelevision under, rather than over (or accurately)reflects the frequency, quantity and place of alcoholconsumption in society. As Robson (1994) notes,a major problem with any content analysis isthat it is never clear what the motives were ofscriptwriters, directors, etc. That is, were theydirectly influenced by advertisers themselves orperhaps by alcohol advertisements, or are theirscripts mainly a consequence of their own, bydefinition, limited and possibly unrepresentativeexperience? As with all non-experimental, cross-sectional approaches, it is inevitably impossible tosay whether alcohol consumption on the televisionencourages more drinking or simply reflects that

Table VL Sex, age and social status of drinking characters

Demographiccharacteristics

Sexmalefemale

Age0-1818-2425-4445-6465 +

Social status

Number

8348

024632321

upper middle class 2middle classworking classunclear

6959

1

Frequency (%)

63.4036.60

18.3048.1017.6016.00

1.552.745

0.8

drinks were verbally offered they were refused34% of the time.

Consequences of drinking

Only three out of 131 characters coded showedany effects from their consumption of alcohol andonly one of these was shown in a negative light,the other two being treated in a light hearted,slightly comic manner. There was no portrayal ofalcohol-related accidents and drink was not shownto have any adverse effects on work ability orconcentration.

Discussion

Television is regarded as a major source of informa-tion about many subjects (including alcohol) for the

526


Dow

nloaded from



which currently occurs. However, the extensiveresearch on the portrayal of men and women intelevision advertisements carried out regularly andin different countries does suggest that it may bepossible to see some clear association betweentelevision portrayals and real life experiences(Fumham and Voli, 1989; Furnham and Bitar,1993).

The analysis of alcohol and drinking in soapoperas indicated that some of these programmescreate and sustain a symbolic environment in whichalcohol consumption is very much the norm, andis particularly associated with sociable behaviour,parties and celebrations (particularly around theChristmas period), and to some degree with affluentlifestyles. In contrast, there is little or no portrayalof the 'negative' side of alcohol consumptionthrough violence, work accidents, family abuse orroad accidents. Few, if any, of the characters areshown to suffer any consequences from their heavyconsumption (cirrhosis, suicide and road accidents)(Hansen, 1984). The problems associated withalcohol abuse tend to relate to a particular storyline(e.g. alcoholism, domestic violence and drink-driving), though it was not detected in any of thesoaps over this period. The pattern of heavy andoften glamorous consumption, unaccompanied byphysical, mental or social damage, will do nothingto highlight the potential dangers of alcohol.Although this pattern of behaviour was not true ofall the soap operas studied here, the fact that itwas most likely to occur in the most watchedserials on British television implies that the implicitmessages about drinking contained in such por-trayals are being exposed to large numbers ofviewers. Indeed, the fact that more than one Britishsoap opera is set in a public house attests to theexposure to a drinking culture in this country.

Overall, the study found that alcohol was presentin approximately 87% of the programmes and that69% of all drinks seen to be consumed werealcoholic. The only programmes in which alcoholwas shown more realistically in amounts, time andplace consumed were the Australian soaps, wherethe majority of drinks seen to be consumed were'soft'. This can be at least in part explained

by the fact that they are aimed at a slightlyyounger audience than say Emmerdale Farm andEastenders, and they are broadcast at midday andbetween 17:30 and 18:30 (times at which schoolchildren are most likely to be watching).

In comparison to Hansen's (1984) and the Pend-leton (1991) study, the evidence suggests that theportrayal of alcohol is still rising, although it couldbe argued that this increase is at least partlydue to the fact that the studies were not strictlycomparable because the three other studies lookedat the incidence of alcohol portrayal in all popularfiction and not specifically soap operas. However,Pendleton et al. (1991) did collect specific data onalcohol presentation in Eastenders and when thiswas examined, alcohol portrayal appeared to haverisen substantially overall. Verbal references toalcohol doubled between 1988 and 1995, with totalreferences (visual and verbal) increased by over60%. The only area in which this was not the casewas in the rate per hour of visual references, butthis can probably be explained in terms of codingdifferences between studies.

In the past, beer has been the most widelyconsumed alcoholic beverage according to popula-tion surveys (PAS, 1995), but in the last few years,wine and spirit consumption has rapidly increased.However, results showed beer is still portrayed farmore frequently than either wine or spirits. Further,portrayal of beer has doubled in the period 1986—95. In addition to this, non-alcoholic (soft) drinkswere shown at almost half the frequency of alco-holic drinks. Alcohol amounts to only 16% of thetotal beverage use in the 'real' world (PAS, 1995).In addition to this, the fact that there was foundto be a fairly substantial number of visual refer-ences to alcoholic brand names, particularly inEastenders, suggests a great deal of 'free advert-ising' to certain drink companies via product place-ments within these sorts of programmes. Nosystematic record was kept of whether brands weregenuine or fictional, but there clearly was evidencethat many of the portrayals were of genuine brands.This distinction should be made in future contentanalytic work.

In agreement with Hansen's work (1984, 1988),

527


Dow

nloaded from


A. Furnham et al.

the majority of alcohol-related acts involved maincharacters from the programmes and it has beensuggested that it is these characters who are themost likely to provide role models to the audience,in particular, the young and impressionable(Wallach et al., 1985). It should be noted, however,that recently many soap operas (Eastenders inparticular) have made a conscious effort to showmain characters under 18 consuming 'soft' drinksand only being served 'soft' drinks when in thepub. The age ranges found were similar to thosefound in other studies, but in contrast to Hansen'swork, quite a large percentage was found to be inthe over 65 years old groups and more interestinglyin the 18-24 age group. This brings into questionthe code governing alcohol advertising on televi-sion, which states that 'no one associated withdrinking in an advertisement should seem to beyounger than about 25' (IBA, 1993). If this isthought to be important in alcohol advertising thenpossibly this restriction should be acknowledgedin television programmes, but more practically itmay be that role models should be encouraged tobehave responsibility. Another difference corres-ponding with Hansen's (1984) findings occuredwith respect to drinking characters on televisionexhibiting showing a 'heavy bias towards thehigher echelons of society*. The characters codedin this study did not show a particular bias tomiddle or upper classes. However, this may wellbe due to the fact that most of the soaps analysedin this study are set in predominantly workingclass areas.

Through close monitoring of alcohol presenta-tion, suggestions could be made to programmeproducers to ensure that a more 'educational' and'real life' picture of alcohol and its associatedproblems be portrayed. However, there has beensome argument against changing exhibiting codesto provide guidelines as to the acceptable presenta-tion of alcohol on television. Some broadcastregulators believe that deliberately generating aform of 'prosocial programming' by adjustingbroadcasting practices to achieve a 'positive social(or moral) practice' would not or have much impacton drinking practices (Towler, 1985; Duval, 1986).

However, as Pendleton et al. (1991) have arguedand this study supports, television portrayals ofalcohol, at present, are neither neutral nor realisticand, if anything, tend to show a bias in favour ofpositive image of alcohol consumption withoutshowing any of the more serious effects that canresult from over-indulgence. The issue is not somuch one of social or moral concern, but is arecognition of the need for accurate informationand sufficient indications of the very real dangersof excessive alcohol use.

Although many of the arguments against theimportance of media content argue that there is,as yet, no clear evidence of any strong 'effect' onviewers in terms of alcohol consumption, the areaof concern should look to the effect these portrayalsmay have in particular on the young (and impres-sionable) and those who are trying to reduce theiralcohol intake (Sobell et al, 1993).

References

Atkin, C. and Block, M. (1981) Content and effects of alcoholadvertising. Report submitted to the Bureau of Alcohol,Tobacco and Firearms, Springfield, VA.

Atkin, C , Hocking, J. and Block, M. (1984) Teenage drinking:does advertising make a difference? Journal ofCommunication, 34, 157-167.

Broadcasters' Audience Research Board (BARB) (1993) AnnualBroadcasting Figures. FTC, London.

Breed, W. and De Foe, J. (1981) The portrayal of the drinkingprocess on prime-time television. Journal of Communication,31, 58-67.

Breed, W., De Foe, J. R. and Wallack, L. (1984) Drinking inthe mass media: a nine year project. Journal of Drug Issues,14, 655-664.

Brown, J. and Walsh-Childers, K. (1994) Effects of media onpersonal and public health. In Bryant, J. and Zillmann, D.(eds), Media Effects. Erlbaum, Hillsdale, NJ, pp 389-415.

Bryant, W. and Zillmann, D. (1994) Media Effects: Advancesin Theory and Research. Erlbaum, Hillsdale, NJ.

Cafiso, J., Goodstadt, M. S., Garlington, W. K. and Sheppard,M. A. (1982) Television portrayal of alcohol and otherbeverages. Journal of Studies on Alcohol, 43, 1233—1243.

Department of Health and Social Security (1981) Preventionand Health: Drinking Sensibly. HMSO, London.

Duval, R. (1986) A fair view on television. Alcohol Concern,2, 13-15.

Feldman, P. (1980) The making and control of offenders. InFeldman, P. and Orford, J. (eds), Psychological Problems:The Social Context. Wiley, New York, pp. 185-218.

Furnham, A. and Voli, V. (1989) Gender stereotyping in Italiantelevision advertisements. Journal of Broadcasting andElectronic Media, 33. 135-185.

528


Dow

nloaded from



Furnham, A. and Bitar, N. (1993) The stereotyped portrayal ofmen and women in British television advertisements. SexRoles, 29,297-310.

Gunter, B., Sancho-Aldridge, J. and Winstone, P. (1994)Tele\'ision: The Public's Vteyv—1993. John Libby, London.

Hanneman, G. and McEwen, W. (1976) The use and abuse ofdrugs; an analysis of mass media content In Ostman, R. E.(ed.), Communication Research and Drug Education. Sage,Beverly Hills, CA, pp. 65-38.

Hansen, A (1984) The Portrayal of Alcohol and Drinking onPrime Time Television. Centre for Mass CommunicationResearch, University of Leicester.

Hansen, A (1988) The Portrayal of Alcohol in PopularTelevision Serials. Centre for Mass Communication Research,University of Leicester.

International Broadcasting Authority (1983) The IBA Code ofAdvertising Standards and Practice. IBA, London.

Kolbe, R. H. and Burnett, M. S. (1991) Content-analysisresearch: an examination of applications with directives forimproving research reliability and objectivity. Journal ofConsumer Research, 18, 243-244.

Lowry, D. T. (1981) Alcohol consumption patterns andconsequences on prime time network TV. JournalismQuarterly, 58, 3-8.

Pendleton, L. L., Smith, C. and Roberts, J. L. (1991) Drinking

on television: a content analysis of recent alcohol portrayal.British Journal of Addiction, 86, 769-774.

Public Attitude Survey (1995) The Drink Pocket Book. NTCPublications, London.

Robson, C. (1994) Real Worid Research. Blackwells, Oxford.Smith, C, Roberts, J. L. and Pendleton, L. L. (1988) Booze

on the box. The portrayal of alcohol on British television: acontent analysis. Health Education Research, 3, 267-272.

Snedecor, G. and Cochran, W. (1989) Statistical Methods, 8thedn. Iowa State University Press, Ames, LA.

Sobell, L. C, Sobell, M. B., Toneatto, T. and Leo, G. I. (1993)Severely dependent alcohol abusers may be vulnerable toalcohol cues in Television programmes. Journal of Studieson Alcohol, 54, 85-91.

Towler, R. (1985) The portrayal of alcohol on television. InInstitute of Alcohol Studies, The Presentation of Alcohol inthe Mass Media. Institute of Alcohol Studies, London, pp.51-58

Wallack, L., Breed, W. and Cruz, J. (1987) Alcohol on prime-time television. Journal of Studies on Alcohol, 48, 33—38.

Wallack, L., Breed, W. and De Foe, J. R. (1985) Alcohol andsoap operas: drinking in the light of day. Journal of DrugEducation, 15, 365-379.

Received on September 20, 1996; accepted February 16, 1997

529


Dow

nloaded from



Dow

nloaded from


A content analysis of alcohol portrayal and drinking in British television soap operas

Documents