1 Smokers' Responses To Anti-Smoking ...

1

Smokers’ Responses To Anti-Smoking Advertisements By Stage Of Change

By

Robert J Donovan* Susan Leivers** Leonard Hannaby***

*Graduate School of Management, University of Western Australia Nedlands, Western Australia Australia 6907 ** Health Department of Western Australia East Perth, Western Australia, 6004 ***Donovan Research West Perth, Western Australia, 6005

Social Marketing Quarterly; vol 5 (2); 1999; 56-65

2

Abstract Prochaska and DiClemente’s stages of change concept has been readily adopted by health promoters and social marketers. These stages have been recommended as a segmentation variable in health promotion and social marketing campaigns, while a recent social marketing text proposes the Prochaska stages as a core element of the social marketing approach. For any marketing segmentation base to be meaningful, it must be shown that the different segments respond differentially to some aspects of the communication and marketing mixes directed at the segments. Hence the utility of the stages of change approach in social marketing is dependent on evidence that individuals in the various stages of change do respond differentially to elements of the social marketing mix. Given that information is the primary component of many social marketing campaigns, and that this information is often carried in the form of paid advertisements or public service announcements (PSAs), it would be useful to determine whether individuals in the different stages of change respond differentially to the same message, and hence require separate communications. In this paper we report results from the Western Australian Health Department’s formative research with respect to which, if any, of three testimonial style ads produced in the USA would be included in the Department’s 1994 ‘Quit!’ campaign. The pre-testing was designed to assess the relative impact of each of the ads on pre-contemplators versus those in all other stages of change. On the pre-test variables of personal relevance, believability, and impact on intentions to quit or cut down the amount smoked, contemplators and those either preparing to take action or already taking action responded significantly and substantially more favorably to all three ads than did pre-contemplators. These results indicate the potential utility of the stages of change as a segmentation variable and suggest that pre-testing of health promoting social marketing communications should be carried out against stage of change segments.

3

Introduction Based on clinical experience with the treatment of addictive behaviours, and smoking behaviour in particular, Prochaska and DiClemente (1983;1984;1986), have proposed a stage model for the adoption of new behaviors. Their transtheoretical model states that people move through a series of stages in adopting a new behavior: • precontemplation: where the individual is not thinking about making a change; • contemplation: where the individual intends to make a change, but not in the immediate future; • preparation or ready for action: where the individual intends to try to make a change in the immediate future and may be making small preparatory changes; • action: where the individual actively attempts the change; and • maintenance: where the individual continues the change behaviour but it requires active or conscious effort to be sustained. Prochaska and DiClemente claim that individuals at different stages of change would have different attitudes, beliefs and motivations with respect to the (desired) new behaviour, and hence different treatment approaches and health communication strategies may be necessary for individuals in the different stages of change. There is some support for these claims over a variety of areas (see Prochaska, Velicer, Rossi et al. 1994 for an analysis of 12 health behaviours), but particularly smoking (Prochaska, Velicer, Di Clemente et al. 1988; DiClemente, Prochaska, Fairhurst et al. 1991), nutrition (Curry, Kristal and Bowen 1992; Campbell, Symons, Demark-Wahnefried et al. 1998), and exercise (Marcus and Owen 1992; Marcus, Rakowski and Rossi 1992; Oman and King 1998; Prochask and Marcus 1994). Prochaska and DiClemente’s stages of change concept has been readily adopted by health promoters and social marketers (e.g., Andreasen 1995; Booth, Macaskill, Owen et al. 1993; Donovan and Owen 1994; Egger, Donovan and Spark 1993; Maibach and Cotton 1995). Donovan and his colleagues have suggested that these stages be used as a segmentation variable in health promotion and social marketing campaigns (Egger, Donovan and Spark 1993; Donovan and Owen 1994), while a recent social marketing text proposes the Prochaska stages as a core element of the social marketing approach (Andreasen 1995). For any marketing segmentation base to be meaningful, it must be shown that the different segments respond differentially to some aspects of the communication and marketing mixes directed at the segments. Hence the utility of the stages of change approach in social marketing is dependent on evidence that individuals in the various stages of change do respond differentially to elements of the social marketing mix. While a number of studies have shown differences between individuals in the different stages of change, much of this has been directed towards the implications for counselling and educational interventions (Prochaska, Velicer, Rossi et al. 1994; Prochaska,Velicer, Di Clemente et al. 1988; DiClemente, Prochaska, Fairhurst et al. 1991; Marcus, Rakowski and Rossi 1992; Marcus, Selby, Niaura et al. 1992; Booth, Macaskill, Owen et al. 1993). However, there is no published evidence that we are aware of that reports differences in response to advertising components of social marketing campaigns. Given that information is the primary component of many social marketing campaigns (Young 1989), and that this information is often carried in the form of paid advertisements or public service announcements (PSAs), it would be

4 useful to determine whether individuals in the different stages of change respond differentially to the same message, and hence require separate communications. In this paper we report part of the results of the Western Australian Health Department’s formative research with respect to which, if any, of three testimonial style ads produced in the USA would be included in the Department’s 1994 ‘Quit!’ campaign. The primary target group for the campaign was defined as non-tertiary educated male and female smokers, aged 25-44 years, and who were contemplators or in the ready for action or action stages. The pretesting was designed to assess the relative impact of each of the ads on pre-contemplators versus those in all other stages of change. From a campaign perspective, if the ads were found to be far more effective with pre-contemplators than with contemplators, then a decision would need to be made whether or not to proceed with these ads or use other ads. Method The advertisements Three television commercials were tested using standard pre-testing procedures (Rossiter and Percy 1997). Two of the ads featured models who had appeared in former cigarette television advertising and who were now suffering from cancer: a male with lung cancer (Marlboro ads) and a female with throat cancer (Lucky Strike ads). In the Quit ads they attribute their cancer to smoking. The third ad showed a former male model for Winston cigarettes advertising criticizing the marketing strategies utilized by the tobacco industry. Sample A convenience sample of N = 264 smokers aged 25-44 took part in the study. Respondents were intercepted by a research agency’s professional interviewers in the downtown area of the state’s capital and screened on the relevant criteria. Those working in the following fields, or who had a close family member working in these fields, were excluded from the study: advertising or market research; and tobacco manufacturing, marketing or retailing. Qualified respondents were invited to the research agency’s ad testing centre and randomly assigned to one of the three ads within the following approximate quotas: two thirds male and one third female; two thirds 25-34 years and one third 35-44 years; and 60% of the sample to be in the contemplation, ready for action or action stages, and 40% to be pre-contemplators. Procedure The testing procedure followed that recommended by Rossiter and Percy (1987; 1997). Respondents were exposed to one only of the ads in groups of three or four persons. Following Krugman (1972), each ad was exposed twice. Post-exposure interviewing was conducted one-on-one by an interviewer using a structured questionnaire. The Questionnaire In the pre-exposure screening questionnaire, respondents were self-classified via their response to the following four point stage of change scale which has been shown to be relatively reliable (Donovan, Jones, Holman et al. 1998): • I am not thinking about giving up smoking (Pre-Contemplation) • I am thinking about trying to give up smoking, but not in the next fortnight (Contemplation) • I am thinking about trying to give up smoking in the next fortnight or so (Ready for action)

5 • I am trying to give up smoking at the moment (Action) The post-exposure questionnaire items were based on standard pre-testing measures (Rossiter and Percy 1997), and consisted of a cognitive response measure; comprehension; emotions experienced while watching the ad (prompted); perceived personal relevance of the ad; credibility of the ad; and impact on considering quitting or cutting down the amount smoked. Also, given the nature of the ads, and particularly the Winston ad, pre and post measures were taken of attitudes to the tobacco industry. Results The proportions in each of the stages of change were: pre-contemplators 42% (n = 111); contemplators 37% (n = 98); ready for action 6% (n = 15); and action 15% (n = 40). Given the small n in the ready for action stage, the ready for action and action stages were combined for all analyses. It should be noted that the three samples did not differ on occupational status and were quota’d to be equivalent with respect to age and gender composition. As ANOVA showed no significant interactions by ad, the results below are presented for all three ads combined. Given the consistent nature of the results across measures, we present the results only for personal relevance; credibility; and impact on considerations of quitting or cutting down. All other variables showed results consistent with these. Personal relevance and credibility measure reactions to the ads per se, whereas impact on considerations of quitting or cutting down measures the ads’ effectiveness in terms of the ads’ communication objective. Personal relevance and credibility can be considered diagnostic variables that mediate ad effectiveness. The specific questions measuring each of the three variables are given below. It was expected (and required from a campaign point of view), that the ads would be more personally relevant and credible, and have more impact on intentions to quit or cut down with those in more advanced stages of change; that is, contemplators would respond more favorably than pre-contemplators, and those in the action and ready for action stages would respond more favorably than contemplators. Personal Relevance: Respondents were asked: “How relevant do you think the ad was to you personally?”, and provided with the five point scale: very, quite, somewhat, not very and not at all. Table 1 presents the percentage results by stage of change with the quite and somewhat categories and the not very and not at all categories combined. The percent in each stage of change nominating the ‘top box’ (‘very relevant’) is shown in Figure 1. Credibility: Respondents were asked: “How believable was the ad?”, and provided with the three point scale: totally, partly and not. Table 2 presents the percentage results by stage of change with the ‘top box’ (‘totally believable’) only responses shown in Figure 2. Impact: Respondents were asked to nominate which of the following five statements applied to them ‘as a result of seeing the ad’: “much or somewhat more likely to seriously consider quitting in the near future, much or somewhat more likely to seriously consider cutting down the amount you smoke in the near future, no more likely to seriously consider quitting or cutting down the amount you smoke”. Table 3 shows the percentage results by stage of change. The much and somewhat categories for quitting and for cutting down respectively,

6 have been combined in Table 3. The ‘top box’ (‘much more likely to seriously consider quitting’) only responses are shown in Figure 3. ANOVA was carried out on each of the three variables for all ads combined. These results are shown in Table 4. All three variables showed a significant difference between stages overall (Relevance, F = 17.95, p = .000; Credibility, F = 6.43, p = .002; Impact, F = 15.96, p = .000), with the later stages of change responding more favorably to the ads than precontemplators. To assess differences between each of the three stages, Chi square analyses were carried out on the top box response versus all other response categories combined for each variable. Chi square analyses showed significant increases in the percent nominating ‘very’ relevant for each stage of change; that is, significantly more of those in the combined action and preparing for action stage than contemplators rated the ads ‘very personally relevant’ (46% vs 29%; Chi square = 4.44, p = .05), and significantly more contemplators than pre-contemplators rated the ads ‘very personally relevant’ (29% vs 13%; Chi square = 8.25, p = .004)(see Figure 1). There was a significant increase in the percent rating the ads ‘totally’ believable from precontemplation to contemplation (60% vs 79%; chi square = 8.06, p = .007), but no difference between the contemplation (79%) and combined action and ready for action stage (78%)(see Figure 2). The Impact measure showed significant increases for each stage of change; that is, significantly more of those in the action and preparing for action stage versus contemplators were ‘very much more likely to consider quitting’ (45% vs 22%; chi square = 8.76, p = .004), as were significantly more contemplators than pre-contemplators (22% vs 11%; chi square = 5.18, p = .03)(see Figure 3). Further demonstration of a stage effect is that, amongst the 74% of those ready for action or in the action stage and the 69% of contemplators who reported an increased likelihood of seriously considering quitting or cutting down in the near future as a result of exposure to the ad (Table 3), the vast majority in each case (76% and 68% respectively) nominated an increased likelihood of quitting rather than cutting down. Conversely, amongst the 51% of pre-contemplators who reported an increased likelihood of seriously considering quitting or cutting down in the near future as a result of exposure to the ad (Table 3), the majority (59%) nominated an increased likelihood of cutting down. Given that attempts to cut down are often the precursor to more serious attempts to quit, these results indicate not only a differential response in cognitive impacts, but a likely differential behavioral response also. Conclusion This study complements previous studies indicating the utility of the stages of change approach for counselling and educational interventions (e.g., Prochaska, Velicer, Rossi et al. 1994), and extends these findings to social marketing communications. This study supports the potential usefulness of Prochaska’s stages of change as a segmentation variable in social marketing interventions as advocated by a number of social marketers (e.g., Andreasen 1995; Donovan and Owen 1994). In this study, pre-contemplators, contemplators and those in the ready for action and action stages differed significantly and substantially in their responses to three anti-tobacco ads. These results suggest that different messages (or message executions) may be necessary for optimal impact with smokers in the different stages of change. For

7 example, our results suggest that ‘call-to-action’ messages to pre-contemplators could stress cutting down, whereas those to later stages would emphasise quitting outright. These results suggest that qualitative formative research to develop anti-tobacco campaigns should ensure that research is conducted separately on smokers in the different stages of change. Similarly, pre-testing of potential ads should be conducted against stage of change segments rather than against ‘all smokers’. Ads that might be very effective against contemplators might be rejected if tested against a sample consisting primarily of pre-contemplators. With respect to implications for pre-testing anti-tobacco advertising, it is noteworthy that substantial majorities of all stages found the ads ‘totally credible’, and that there was no difference between contemplators and those in the ready for action and action stages on this measure. Hence this measure may be a necessary, but not sufficient, indicator of an ad’s potential effectiveness. Our results suggest that personal relevance - given credibility - is a more valid indicator of potential effectiveness. Finally, this study’s findings are somewhat constrained by the need to collapse the ready for action and action stages given insufficient numbers to analyse these stages separately. Future studies would benefit by ensuring sufficient numbers in each of these stages so as to allow comparisons between all stages. Implications for Social Marketing The stages of change concept has been advanced as a core concept for developing social marketing strategies. It has been specifically suggested as a segmentation variable for defining target audiences in terms of the extent to which they appear receptive to making some behavior change. Previous research has shown that people in the different stages differ in beliefs and attitudes with respect to various health issues, and that different counselling or other interventions are more or less appropriate for people in the different stages. This study complements previous studies by indicating the potential utility of the stages of change concept for mass media communications. It also indicates that formative message strategy developmental research and pre-testing should be carried out against people in the different stages of change, rather than simply ‘all members’ of the target audience.

8 References Andreasen,A.R. (1995). Marketing social change: Changing behavior to promote health, social development, and the environment. San Francisco, CA: Jossey-Bass. Booth,M.L., Macaskill,P., Owen,N., Oldenburg,B., Marcus,B.H. and Bauman,A. (1993). Population prevalence and correlates of stages of change in physical activity. Health Education Quarterly, 20, 431-440. Campbell,M.K., Symons,M., Demark-Wahnefried,W., Polhamus,B., Bernhardt,J.M., McClelland,J.W. and Washington,C. (1998). Stages of change and psychological correlates of fruit and vegetable consumption among rural African-American church members. American Journal of Health Promotion, 12(3), 185-191. Curry,S.J., Kristal,A.R. and Bowen,D.J. (1992). An application of the stage model of behaviour change to dietary fat reduction. Health Education Research, 7, 97-105. DiClemente,C.C., Prochaska,J.O. Fairhurst,S., Velicer,W.F., Velasquez,M.M. and Rossi,J.S. (1991). The process of smoking cessation: an analysis of precontemplation, contemplation and preparation stages of change. Journal of Consulting and Clinical Psychology, 59, 295-304. Donovan,R.J., Jones,S., Holman,C.D.J. and Corti,B. (1998). Assessing the reliability of a stage of change scale. Health Education Research, 13(2), 101-107. Donovan,R.J. and Owen,N. (1994). Social marketing and population interventions. In Dishman,R.K. (ed). Advances in exercise adherence. Champaign, IL: Human Kinetics. Egger,G., Donovan,R.J. and Spark,R. (1993). Health and the media: Principles and practices for health promotion. Sydney: McGraw-Hill. Krugman, H.E. (1972). Why Three Exposures May be Enough. Journal of Advertising Research, 12(6), 11-14. Maibach,E.W. and Cotton,D. (1995). Moving people to behavior change: A staged social cognitive approach to message design. In Maibach,E. and Parrott,R.L. (eds). Designing Health Messages. London: Sage. Marcus, B.H. and Owen,N. (1992). Motivational readiness, self-efficacy and decisionmaking for exercise. Journal of Applied Social Psychology, 22, 3-16. Marcus,B.H., Rakowski,W. and Rossi,J.S. (1992). Assessing motivational readiness and decision making for exercise. Health Psychology, 11, 257-261. Marcus,B.H., Selby,V., Niaura,R.S. and Rossi,J.S. (1992). Self-efficacy and the stages of exercise behaviour change. Research Quarterly for Exercise and Sport, 63, 60-66. Oman,R.F. and King,A.C. (1998). Predicting the adoption and maintenance of exercise participation using self-efficacy and previous exercise participation rates. American Journal of Health Promotion, 12(3), 154-161.

9

Prochaska,J.O. and DiClemente,C.C. (1983). Stages and processes of self-change of smoking: toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390-395. Prochaska,J.O. and DiClemente,C.C. (1984). The transtheoretical approach: Crossing the traditional boundaries of therapy. Dow-Jones/Irwin, Illinios. Prochaska,J.O. and DiClemente,C.C. (1986). Toward a comprehensive model of change. In Miller,W.R. and Heather,N. (eds). Treating addictive behaviours: Processes of change. Plenum Press, New York. Prochaska,J.O. and Marcus,B.H. (1994). The transtheoretical model: Applications to exercise. In Dishman,R.K. (ed). Advances in exercise adherence. Champaign, IL: Human Kinetics. Prochaska,J.O., Velicer,W.F., DiClemente,C.C. and Fava,J. (1988). Measuring processes of change: applications to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520-528. Prochaska,J.O., Velicer,W.F., Rossi,J.S., Goldstein,M.G., Marcus,B.H., Rakowski,W., Fiore,C., Harlow,L.L., Redding,C.A., Rosenbloom,D. and Rossi,S.R. (1994). Stages of change and decisional balance for 12 problem behaviours. Health Psychology, 13, 39-46. Rossiter,J.R. and Percy,L. (1997). Advertising communications and promotion management. New York: McGraw-Hill. Rossiter,J.R. and Percy,L. (1987). Advertising and promotion management. New York: McGraw-Hill. Young,E. (1989). ‘Social marketing in the information era’. American Marketing Association Conference, Social Marketing for the 1990s, Ottowa, Canada.

1

9

Table 1: Perceived Personal Relevance of Ads % Pre% Contemplators Contemplators (n=98) (n=111)

% Ready for Action/Action (n=55)

Very relevant

13

29

46

Relevant

44

53

36

Not relevant

43

18

18

TOTAL

100

100

100

10

Table 2: Perceived Believability of Ads % Pre% Contemplators Contemplators (n=98) (n=111)


Totally believable

60

79

78

Partly believable

30

20

16

Not believable

10

1

6

TOTAL

100

100

100

11

Table 3: Impact of Ads on Likelihood of Quitting or Cutting Down % Pre% Contemplators Contemplators (n=98) (n=111)


More likely to seriously consider quitting in the near future

21

47

56

More likely to seriously consider cutting down in the near future

30

22

18

No more likely to consider quitting/cutting down

49

31

26

TOTAL

100

100

100

12

Table 4: ANOVA Summary Results

Mean

PreContemplators Contemplators (n=98) (n=111)

Ready for Action/Action (n=55)

F value

p value

Relevance*

3.29

2.41

2.22

17.95

.000

Credibility*

1.5

1.22

1.27

6.43

.002

Impact*

3.86

2.99

2.55

15.96

.000

* Lower scores indicate more favorable response.

12

Figure 1: Perceived Personal Relevance of Ads By Stage of Change 50 45 40 35 30 % Very Relevant 25 (‘Top box’) 20 15 10 5 0

PreReady for Contemplators Contemplators Action/Action

13

Figure 2: Perceived Believability of Ads By Stage of Change 100 90 80 70 % Totally 60 Believable 50 (‘Top box’) 40 30 20 10 0


14

Figure 3: Impact of Ads on Likelihood of Quitting or Cutting Down By Stage of Change 50 45 40 % Much More 35 Likely to Quit 30 (‘Top box’) 25 20 15 10 5 0


15

Rob Donovan, PhD, is Co-Director of the Health Promotion Evaluation Unit and a Professorial Fellow in marketing in the Graduate School of Management at the University of Western Australia. Susan Leivers is Co-ordinator Research and Evaluation in the Health Promotion Services Branch in the Health Department of Western Australia. Leonard Hannaby is Systems and Data Manager at NFO - Donovan Research, a social and marketing research company based in Perth, Western Australia.