The Effects of Scheduling Format and Booster Sessions ... - CiteSeerX

Journalof Behavioral Medicine, Vol. 6, No. 4, 1983

The Effects of Scheduling Format and Booster Sessions on a Broad-spectrum Psychosocial Approach to Smoking Prevention Gilbert J. Botvin,' Nancy L. Renick,l and Eli Baker1 Accepted forpublication: February 27, 1983

The effectiveness of a fifteen session psychosocial smoking prevention strategy was tested on 902 seventh graders from seven junior high schools in suburban New York over 2 years. The prevention program was implemented by regular classroom teachers and consisted of a cognitive component dealing with the immediate consequences of cigarette smoking, a decision-making component, a relaxation-training component, a social skills training component, and a self-improvement component. In addition to testing the overall effectiveness of this approach, the relative efficacy of two duferent scheduling formats was compared and the extent to which "booster" sessions conducted during the year after completion of the program helped to maintain reductions in new smoking was also examined. Results indicated that the prevention program was able to reduce new cigarette smoking by 50% at the end of the first year and by 55% at the end of the second year for the intensive formal condition. New regular cigarette smoking was reduced by 87% in the second year for the students in the booster condition. Significant changes consistent with nonsmoking were also evident on several cognitive, attitudinal, and personality variables. KEY WORDS: attitudes; booster sessions; cigarette smoking; cognition; personality; prevention.

This work was supported by a grant to the senior author from the National Institute of Child Health and Human Development (HD 14678-03). 'Cornell University Medical College, 41 1 East 69th Street, New York, New York 10021.

359 0160-7715/83/12M)-035%03.00/0

O 1983 Plenum Publishing Corporation

Botvin, Renick, and Baker

INTRODUCTION Despite an increased public awareness of the health hazards of cigarette smoking and a decline in the adult prevalence rates, cigarette smoking among children and adolescents continues to be a significant problem (U.S. Public Health Service, 1979). Traditional smoking education programs have attempted to deter smoking onset by making students aware of the adverse health consequences of cigarette smoking. The available evidence now indicates that while traditional smoking education programs may be able to increase smoking-related health knowledge and facilitate the development of antismoking attitudes, they have little impact on actual smoking behavior (Evans el al., 1979; Thompson, 1979). Recently, there has been a shift toward the development of prevention strategies which focus on the psychosocial factors promoting the onset of cigarette smoking (Botvin et al., 1980; Evans et a/., 1978; Hurd et al., 1980; McAlister et al., 1979). Although all of these strategies have their roots in social learning theory (Bandura, 1969), they differ in terms of emphasis and mode of implementation. Nearly all of these approaches place primary emphasis on familiarizing students with the various social pressures to smoke and teaching them techniques for resisting these pressures, particularly pressure from their peers. Intervention programs utilizing this type of prevention strategy have been implemented by means of videotape and class discussion (Evans et al., 1978, 1981; Hurd et al., 1980) and by means of peer-led discussion and role playing (McAlister el al., 1979, 1980). Botvin and his collegues, on the other hand, have developed and tested a psychosocial smoking prevention strategy which is somewhat broader in its focus (Botvin, 1982). While techniques for dealing with social pressures to smoking are included, they are presented within the context of a program designed to improve social competence and to facilitate the acquisition of general coping skills. This approach has been successfully implemented by outside health professionals (Botvin et al., 1980; Botvin and Eng, 1980b) and older peer leaders (Botvin and Eng, 1982). The development of these psychosocial smoking prevention strategies clearly represents an important advance in the search for an effective approach to the prevention of adolescent cigarette smoking. These strategies, on the whole, have been found to be capable of producing initial reductions in smoking behaviors of about 50%. Moreover, longer-term results look promising (Botvin and Eng, 1982; Evans et al., 1981; McAlister el al., 1980; Murray et al., 1979). However, these strategies of necessity have been implemented under conditions not typical of most classrooms. Thus, while the existing evidence now supports the efficacy of psycho-

Booster Sessions and Smoking Prevention

361

social prevention strategies when tested under well-controlled conditions, it is unclear to what extent they might succeed under more typical classroom conditions. The main purpose of the present study, therefore, was to provide a realistic test of the effectiveness of one such psychosocial smoking prevention strategy when implemented by regular classroom teachers under fairly typical classroom conditions. Additionally, the present study was designed to compare the relative efficacy of two different scheduling formats, to examine the extent to which reductions in new smoking at the end of the program resulted in reductions in new regular smoking one year later, and to test the extent to which maintenance ("booster") sessions conducted during the year after completion of the program helped to maintain the preventive gains achieved at the end of the program. METHODS Subjects

Seventh-grade students from seven suburban New York schools were included in the present study. The students in these schools were predominently White (91%) and were from middle- to upper-middle-class families. Table I presents the baseline smoking rates and pattern of smoking Table 1. Baseline Smoking Rates and Patterns Broken Down by School and Condition

Never smoked

Tried but auit

Currcnt occasional smoking

Current regular smoking

Experimental (E,) School No. 5 School No. 7 Combined

66 86 80

18 12 I4

10 3 5

6 0 2

Experimental (E,) School No. 1 School No. 4 Combined

79

86 83

8 11 10

10 3 6

3 1 2

Control School No. 2 School No. 3 School No. 6 Combined

85 85 75 82

8 10 15 11

6 4 7 6

2 I 2 2

Total Experimental Control

81 82

12 II

5 6

2 2

Botvin, Renick, and Baker

362

for the schools assigned to the experimental and control conditions. Although, some degree of interschool variability is evident in terms of both occasional and regular smoking, the combined rates for each condition show little intercondition variability. Pretest and posttest data were collected from 902 students. Of these, 831 (92%) were classified as nonsmokers at the time of the pretest. This group comprised the nonsmoking cohort which was the focus of attention over the course of the intervention and follow-up phases of the study.

Design The seven schools in the study were randomly assigned to the following conditions: ( 1 ) LST Smoking Prevention Program, utilizing an integrated weekly scheduling format (E,); (2) LST Smoking Prevention Program, utilizing an intensive minicourse scheduling format (E,); and (3) control. Two schools were assigned to each of the two experimental conditions, and three schools were assigned to the control condition. Data were collected from all groups at the pretest, posttest, and 1-year follow-up. The smoking prevention program was implemented between the pretest and the posttest for both experimental groups. In addition, students from one school (E2B) participated in an eight-session booster program which was implemented in the fall of the second year of the study between the posttest and the 1-year follow-up.

Procedure All of the students in the study were pretested by questionnaire for self-reported smoking status and on several cognitive, attitudinal, and personality variables believed to be related to the initiation of cigarette smoking. Moreover, saliva samples were collected immediately prior to the completion of the self-report section of the questionnaire in a variant of the "bogus pipeline" procedure (Evans et al., 1977). Students in the two experimental conditions participated in the 15-session Life Skills Training (LST) Smoking Prevention Program which was implemented by regular classroom teachers. Approximately 4 months after the pretest, all students were posttested by questionnaire and saliva samples were collected. Some students received an additional eight-session "booster" program in the second year of the project. One pear after the posttest, students once again were tested by means of questionnaire and saliva samples were collected.

Booster Sessions and Smoking Prevention

363

Experimental Conditions. Students in both experimental conditions participated in a 15-session smoking prevention program. Each session was one class period (about 1 hr in length). The only difference between the two experimental conditions was with respect to the implementation schedule. In one of the experimental conditions (E,), the smoking prevention program was integrated into the regular science or health curriculum and was conducted at a rate of one session per week. In the other experimental condition (Ez), the smoking prevention program was scheduled as an intensive minicourse or curriculum module in which the LST Program was conducted on consecutive class days over the course of about 1 month. The main emphasis of the LST Prevention Program was on the acquisition of basic life skills and the improvement of general personal competence. Sessions were included which dealt with the effects of cigarette smoking (particularly the more immediate physiological effects), self-image and self-improvement, decision making, techniques used by advertisers, coping with anxiety, communication skills, social skills, and assertiveness (including techniques for resisting peer pressure to smoke). Table I1 provides a brief description of each session. A detailed description of the curriculum including individual lesson plans is provided elsewhere (Botvin and Eng, 1980a). The material covered in the program was taught through the use of a variety of techniques including instruction, group discussion, modeling, and behavior rehearsal. The LST Program was implemented by regular classroom teachers under the general supervision of members of the project staff. Prior to the implementation of the prevention program, all teachers attended a I-day orientation workshop in order to become familiar with the LST curriculum. All teachers were provided with a Teacher's Manual (Botvin and Eng, 1980a) which contained detailed lesson plans for the entire program. In addition, members of the project staff were available to consult with the teachers whenever necessary, Teachers were monitored periodically by the project staff to assess the extent to which they were following the intervention protocol and to assure that the LST Program was being properly implemented. Control Condition. The students in the control condition were not provided with any special smoking prevention activities beyond the traditional smoking education efforts mandated by law in New York State. The data collection procedures followed for the students in the control group were the same as those used for the experimental group. Control students were told that they were participating in a survey of teenage cigarette smoking in order to provide an explanation for the data collection.

Botvin, Renick, and Baker

364

Table 11. Description of Life Skills Training Program Session

Material covered

Orientation

General introduction to the program, saliva collection and administration of pretest questionnaire, overview of forthcoming sessions

Smoking: myths and realities

Common attitudes and beliefs about smoking; prevalence of smoking, reasons for and against smoking, the process of becoming an addicted smoker, and the decreasing social acceptability of smoking

Smoking and biofeedback

Effect of smoking on carbon monoxide levels and heart rate are demonstrated using ecolyzer and cardiotachometer

Self-image and selfimprovement

Self-image and how it is formed, the relationship between self-image and behavior, the importance of a positive selfimage, and ways of improving self-image

Decision-making and independent thinking

A general decision-making strategy, decision-making and sources of influence affecting decisions, resisting persuasive tactics, and the importance of independent thinking

Advertising techniques

Use and function of advertising, ad techniques, identifying techniques used in cigarette advertising and how they are designed to affect consumer behavior, alternative ways of responding lo these ads

Coping with anxiety

Situations causing anxiety, demonstration and practice of techniques for coping with anxiety

Communication skills

Verbal and nonverbal communication, techniques for avoiding misunderstandings, basic conversational skills, giving and receiving compliments, making introductions, etc.

Social skills A

Overcoming shyness, initiating social contacts, giving and receiving compliments, basic conversational skills

Social skills B

Boy-girl relationships, nature of attraction, conversing with the opposite sex, asking someone out for a date

Assertiveness

Situations calling for assertiveness, reasons for not being assertive, verbal and nonverbal assertive skills, resisting peer pressures to smoke

Conclusion

Brief review, conclusions, saliva collection, posttest questionnaire

Measures of Smoking Status. Smoking status was determined by means of three self-report measures: a monthly recall measure, a 7-day (weekly) recall measure, and a 24-hr (daily) recall measure. The monthly recall measure provides a means of detecting even relatively infrequent smoking, whereas the weekly and daily measures provide a means of detecting more regular smoking. Because previous research has raised questions concerning the validity of self-reported smoking status in junior high students (Evans et al., 1977; Hurd et al., 1980), steps were taken to ensure the quality of the self-reported data. First, students' 1.D. codes were

Booster Sessions and Smoking Prevention

365

utilized rather than names to facilitate following each student throughout the course of the study while at the same time emphasizing the confidential nature of the questionnaire. Second, saliva samples were collected immediately prior to the section of the questionnaire dealing with selfreported smoking status. Such a procedure has been found to be capable of increasing the accuracy of self-reports (Evans et al., 1977). Cognitive Measures. In order to measure changes in knowledge resulting from the intervention program, 20 true-false items were included in the questionnaire. Half of these items were designed to measure changes in knowledge relating t o cigarette smoking (e.g., prevalence of smoking among adults and adolescents, immediate physiological effects of smoking). The other half were designed to measure changes in psychosocial knowledge (e.g., knowledge concerning techniques used by advertisers, knowledge concerning specific interpersonal skills). Test-retest reliability estimates for both smoking-related knowledge (0.76) and psychosocial knowledge (0.75) were good, suggesting that both measures are reliable. Moreover, based on the use of these items in our previous research, they have been found to have a high degree of construct validity. Attitudinal Measures. Students' attitudes about cigarette smoking, the characteristics of smokers, the social benefits of smoking, the health consequences of smoking, etc., were assessed by means of a 20-item scale derived from the Teenager's Self Test: Cigarette Smoking (U.S. Public Health Service, 1974). The items contained in the attitude scale used in this study were selected from a longer list of 64 items based on their age appropriateness and their ability to discriminate between smokers and nonsmokers. Personality Measures. A number of personality variables were assessed using several different scales. Assertiveness was measured using a shortened (20-item) version of the Assertion Inventory (Gambrill and Richey, 1975). Locus of control was measured by means of an empirically shortened (six-item) version of Rotter's I/E Scale (Schlegel and DiTecco, 1978). Social anxiety was measured by means of seven situation-specific items relating to common social situations which might produce anxiety (e.g., giving and receiving compliments, expressing feelings, initiating conversations). Self-esteem was measured using self-ratings in response to 10 descriptive adjectives (e.g., smart, popular, good-looking) in a manner typical of most measures of self-esteem (Wells and Marwell, 1976). Selfconfidence and self-satisfaction were measured using three descriptive statements each (e.g., "I am able to handle difficult situations," "I am generally satisfied with myself'?. Smoking influenceability was measured using three items concerning the respondent's probability of yielding to social pressure to smoke. General influenceability was measured using five items concerning the extent to which the respondent is influenced by

Botvin, Renick, and Baker

366

others when making an important decision. With the exception of locus of control, all of the personality variables were measured using five point Likert-type scales. Furthermore, all of these items have been used in our previous research and have test-retest reliabilities ranging from 0.66 to 0.78 (Botvin and Eng, 1982).

RESULTS Smoking Behavior: Year 1

Changes from Nonsmoking to Smoking. Table 111 presents the distribution of pretest nonsmokers who began smoking prior to the posttest based on the monthly, weekly, and daily recall measure of smoking behavior. Comparison of the experimental and control groups indicated onset rates of 6 and 13%, respectively, using the monthly measure, 4 and 6% using the weekly measure, and 1 and 3% using the daily measure. A chi-square analysis of the proportion of pretest nonsmokers who began smoking prior to the posttest was significant using the monthly measure [x2(1) = 10.05, P < 0.0011 but was not significant when more frequent smoking was considered using either the weekly measure or the daily measure. Comparison of the relative effectiveness of the integrated weekly intervention format (El) and the intensive minicourse format (Ez) inTable 111. Number of Pretest Nonsmokers Who Reported Smoking at the Posttest Based on Monthly, Weekly, and Daily Recall Measures: Comparison of E,, E,, and Control Conditions (N= 9021

Measure of smoking status and condition Monthly measure Experimental (E,) Experimental (E,) Control Weekly measure Experimental (El) Experimental (E2) Control Daily measure Experimental (EJ Experimental (EJ Control

Pretest nonsmokers

Posttest nonsmokers

Posttest smokers

Onset rate (%)

Booster Sessions and Smoking Prevention

367

dicated onset rates of 7 and 6070, respectively, for the monthly measure, 4 and 3% for the weekly measure, and 1% each for the daily measure. A series of chi-square analyses was computed to determine more specifically the effectiveness of each of the two intervention formats when compared with the control group. Both the E, condition [x2(1) = 6.84, P < 0.021 and the Ez condition [x2(1) = 4.92, P < 0.021 produced significantly fewer new smokers when compared to the control group in terms of the monthly recall measure. However, no significant differences were found for either the E, condition or the Ez condition using the two measures of more regular smoking. Overall, it appears that both conditions were equally effective in preventing the onset of new smoking when using the most general (monthly) measure of smoking behavior. Other Changes in Smoking Behavior. Table IV presents the distribution of pretest occasional smokers (individuals reporting having smoked in the past month but not in the past week) and regular smokers (individuals reporting having smoked in the past week or day) in the experimental and control groups according to their smoking status at the time of the posttest. Because of the small number of pretest smokers in each of the posttest categories, both experimental groups were combined for the purpose of these comparisons. Fewer pretest occasional smokers began smoking more regularly at the time of the posttest in the combined experimental group (25%) than in the control group (33%), while more of the pretest occasional smokers in the experimental group (60%) quit smoking than in the control group (50%). Similarly, more of the pretest regular smokers in the combined experimental group either reported less frequent smoking than those in the control group (27 vs 15%) or reported having quit smoking (36 vs 23%). Although all of these differences favored the experimental group, none was statistically significant.

Table IV. Posttest Smoking Status of Pretest Occasional and Regular Smokers Posttest smoking status Pretest smoking status and condition Occasional Experimental Control Regular Experimental Control

Occasional

Regular Quitter -

N

(To)

N

(TO)

N

(%)

3 4

(15) (17)

5

(25) (33)

12

12

(60) (50)

3 2

(27) (15)

4

(36) (62)

4 3

(36) (23)

8

8

Botvin, Renick, and Baker

368

Smoking Behavior: Year 2 Changes from Nonsmoking to Smoking. Table V presents the distribution of pretest nonsmokers with respect t o their smoking status at the 1-year follow-up. The onset rates for the experimental and control groups were 15 and 22070, respectively, using the monthly measure, 8 and 15% using the weekly measure, and 6 and 11% using the daily measure. Chi-square analyses of the proportion of pretest nonsmokers who began smoking prior to the 1-year follow-up were significant using the monthly measure [x2(1) = 3.93, P < 0.051, the weekly measure [x2(I) = 8.52, P < 0.0041, and the daily measure [x2(1) = 4.58, P < 0.031. Examination of the two experimental conditions individually indicates that the onset rates for the E, and E1 conditions were 19 and lo%, respectiley, using the monthly measure, 11 and 4% using the weekly measure, and 9 and 3% using the daily measure. Comparisons of each of the two treatment conditions with the controls indicate that there were significantly fewer new smokers in the Ez condition than in the control condition using the monthly measure [xZ(l) = 7.30, P < 0.0051, weekly measure [x2(1) = 11.19, P < 0.00081, and daily measure [x2(1) = 7.12, P < 0.0081. However, there were no significant differences in the proportion of new smokers between the El and the control conditions. Chi-square com-

Table V. Number of Pretest Nonsmokers Who Reported Smoking at the OneYear Follow-Up Based on Monthly, Weekly, and Daily Recall Measures: Comparison of El, E,, and Control Conditions

Measure of smoking status and condition

Pretest nonsmokers

Follow-up nonsmokers

Follow-up smokers

Onset rate (%)

Monthly measure Experimental (El) Experimental (E,) ~ontr&

159 125 32 1

128 112 25 1

31 13".~ 70

19 10 22

Weekly measure Experimental (E,) Experimental (E,) Control

167 127 339

149 122 287

18 5b.e 52

11 4 15

Daily measure Experimental (E,) Experimental (E2)

163 127

149 122

14 4c.1

9 3

Note. Superscripts a, b, and c refer to comparisons of E2 vs controls for each measure of smoking status. Superscripts d, e, and f refer to comparisons of El vs E2 for each measure of smoking status. Only significant or nearly significant results are reported. (a) P < 0.005. (b) P < 0.001. (c) P < 0.008. (d) P < 0.04. (e) P 0.03. (0 P ' < 0.06.