Development and preliminary evaluation of a measure of support provided to a smoker among young adults

Addictive Behaviors 30 (2005) 1351 – 1369

Development and preliminary evaluation of a measure of support provided to a smoker among young adults Janet L. Thomasa,T, Christi A. Pattenb, Paul A. Deckerc, Ivana T. Croghand, Margaret L. Cowlese, Carrie A. Bronarsf, Liza M. Nirellic, Kenneth P. Offordc a

Department of Preventive Medicine and Public Health, Mail Stop 1008; University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, United States b Department of Psychiatry and Psychology, Mayo Clinic, Rochester, Minnesota, United States c Division of Biostatistics, Mayo Clinic, Rochester, Minnesota, United States d Nicotine Research Program, Mayo Clinic, Rochester, Minnesota, United States e Louisiana State University, Baton Rouge, Louisiana, United States f University of Kansas, Lawrence, Kansas, United States

Abstract Most studies indicate a positive association between social support and smoking cessation. However, clinic-based interventions to increase support for stopping smoking have had limited success. Prior research has emphasized the smoker’s perceptions of support received for smoking cessation while less attention has focused on support persons’ reports of supportive behaviors provided to a smoker. This study examined select psychometric properties of the Support Provided Measure (SPM), a self-report questionnaire designed by the investigative team to assess supportive behaviors provided to a smoker. The SPM was administered to a college sample (N = 771; 67% female) of young adults, aged 18 to 24 years, who reported knowing a smoker whom they thought should quit smoking. Results indicate that, in this sample, the SPM has a two-factor structure with good internal consistency reliability (Cronbach’s alpha = 0.77) and appears to assess a wide range of individual differences in the provision of support. Demographic correlates associated with SPM scores are described and suggestions for future research are offered. D 2005 Elsevier Ltd. All rights reserved. Keywords: Social support; Smoking; Measurement; College students

T Corresponding author. Tel.: +1 913 588 1568; fax: +1 913 588 2780. E-mail address: [email protected] (J.L. Thomas). 0306-4603/$ - see front matter D 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2005.01.013

1352

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1. Introduction Cigarette smoking is the single most preventable cause of morbidity and mortality (Centers for Disease Control and Prevention, 2002). Clinic-based treatments produce the highest smoking cessation rates but also the lowest population coverage, with less than 5% of smokers taking advantage of these programs (Lichtenstein & Hollis, 1992). Interventions provided by family and friends in the smoker’s natural environment might be an important means of reaching the majority of smokers who do not seek treatment for smoking cessation (Herzog, Abrams, Emmons, & Linnan, 2000; Smith & Meyers, 2004). However, few measures have been developed to assess supportive behaviors provided by others to a smoker. Thus, this study reports on the development and select psychometric properties of the Support Provided Measure, a self-administered questionnaire of smoking specific behaviors directed at helping someone to quit smoking. The importance of extra-treatment support in smoking cessation is documented in the Clinical Practice Guideline on Treating Tobacco Use and Dependence (Fiore et al., 2000). Numerous studies indicate that the type of support that smokers report they receive during quit attempts influences the likelihood of successful smoking abstinence (Cohen & Lichtenstein, 1990; Hanson, Isacsson, Janzon, & Lindell, 1990; Roski, Schmid, & Lando, 1996). Behaviors perceived by the smoker as supportive and that facilitate abstinence include praise and encouragement, the provision of information, showing empathy and concern, tolerating moodiness, offering general problem solving advice and minimizing stress by avoiding interpersonal conflict and by taking over some of the smoker’s responsibilities (Fisher, 1997). Despite these findings, adding a support person component to clinic-based interventions targeting smokers has not been consistently effective (for reviews, see Lichtenstein, Glasgow & Abrams, 1986; Palmer, Baucom & McBride, 2000; Park, Tudiver, Schultz, & Cmpbell, 2004). However, efforts to increase natural support networks within the context of self-help or community-based smoking cessation interventions have been shown to be associated with higher abstinence rates (e.g., Carlson, Goodey, Bennett, Taenzer, & Koopmans, 2002; Gruder, Mermelstein, Kirkendol, & Hedeker, 1993; Pirie, Rooney, Pechacek, Lando, & Schmid, 1997). Thus, the challenge for the field is to determine how to optimize the role of social support in smoking cessation interventions. It is unclear whether the inconsistencies documented in the literature stem from the overall lack of construct refinement of supportive behaviors for smoking cessation (Breteler, Schotborg, & Schippers, 1996), the lack of information about what types of behaviors a support provider is offering (Cohen & Lichtenstein, 1990), or because many studies focused solely on the spouse as the support provider (Collins, Emont, & Zywiak, 1990). It is our contention that each of these limitations need to be addressed. Construct refinement can only come from further exploration of the specific behaviors that a support provider engages in. An important methodological limitation of most prior research is that the smokers’ perceptions of support received have been the sole indicator of support for cessation. Spouses or others who interact with a smoker have rarely been asked about the amount of support they provide (for exception, see Lichtenstein, Andrews, Barckley, Akers, & Severson, 2002;

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1353

Pollak et al., 2001). Relying on one perspective alone may not accurately represent the amount of support given. Prior studies indicate that reports of shared experiences may not be highly correlated and could be influenced by the individuals’ satisfaction with the relationship and their current mood state (Elwood & Jacobson, 1988). This emphasizes the importance of clarifying the intended or actual, as opposed to the perceived supportive behaviors, associated with successful quit attempts. Only then can we adequately develop programs to teach these behaviors to concerned others, design ways to involve them in treatment or follow-up, and identify smokers in need of additional social support. The most commonly used measure of support received for smoking cessation is the Partner Interaction Questionnaire (PIQ; Mermelstein, Lichtenstein, & McIntyre, 1983). Designed to be completed by smokers who report high levels of readiness to quit (i.e., enrolled in a smoking cessation program) and who are married or living with a partner, the PIQ assesses the occurrence and subjective impact of smoking-related interactions between partners during a smoking quit attempt. A limitation of the PIQ is that it does not assess support received among the majority of smokers in the population who report lower levels of readiness to quit (i.e., not engaged in active efforts to quit) (Boyle, O’Connor, Pronk, & Tan, 2000). Further, the PIQ is not applicable to support persons other than the smoker’s spouse or partner. One of the goals of this study was to examine support provided to a smoker in a sample of young adult college students. Most studies examining social support interventions have utilized spouses or other adults to assist a smoker. Less attention has focused on engaging younger individuals to assist a smoker in quitting (Albrect, Payne, Stone, & Reynolds, 1998; Prince, 1995). Nonetheless, some research has shown that younger individuals are interested in helping others with smoking cessation and other health behavior change (Black, Tobler, & Sciacca, 1998; Patten, Lopez, Thomas, Offord, & Decker, 2004). In a preliminary study, (Patten, Offord et al., 2004; Thomas, Patten, Offord, & Decker, 2004), we developed the Support Interview, a measure of supportive behaviors provided to a smoker which was administered using an interview format as part of a clinic-based intervention for support persons. The Support Interview was found to have satisfactory internal consistency reliability (alpha = 0.78 at baseline and 0.84 at week 6) in a sample of 60 adult support persons. However, this measure focused on questions addressing behaviors that were a focus of an intervention, limiting external validity. The measure was therefore revised (Thomas et al., 2004), adapting it to be administered in a self-report format and adding additional behaviors found to be important in smoking cessation efforts. The revised measure, entitled the Support Provided Measure, was designed to assess support provided by a concerned other to a smoker at any level of readiness to quit smoking. The current investigation describes the development and select psychometric properties of the SPM, including item endorsement rates, internal consistency reliability of individual items, item-total score analysis, and the factor structure of the items when completed by a college sample of young adults. Further, we explored potential demographic and tobacco use characteristics associated with the provision of support in this sample. No existing literature was available to guide hypotheses regarding factors that would potentially correlate with reports of support provided to a smoker.

1354

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

2. Research design and methods 2.1. Development of the support provided measure 2.1.1. Content domains and item development Based upon Cohen’s main effects model of supportive interactions (Lakey & Cohen, 2000), a total of 29 items were generated for the Support Provided Measure (SPM). We attempted to identify 5 key questions thought to tap each of five smoking-specific support domains. However, we increased the number of items assigned to assess problem-solving behaviors to 8 and directive/motivating behaviors to 6. Given the exploratory nature of this research and our intent to assess multiple tangible behaviors that may be associated with quitting in the smoker, we balanced the number of items (i.e., subject burden, length of the measure) with sufficient coverage of each domain. The five content domains were: directive/motivating behaviors (items 2: bdiscussed quittingQ, 6: bmodeled exercising behaviorsQ, 7: bmodeled healthy eatingQ, 10: bdiscouraged smoking in certain areasQ, 14: bengaged in a smoke-free activityQ, and 21: bexpressed concern about smoker’s healthQ); punishing behaviors (reverse-scored) (items 3: bnagged or preachedQ, 23: battempted to hide cigarettesQ, 24: bpunished or withheld rewardsQ, 25: bcriticized or blamedQ, and 29: bencouraged smoker to smokeQ); problem solving behaviors (items 1: basked smoker if willing to discuss smokingQ, 5: bgave smoker info on smokingQ, 9: bhelp smoker come up with ideasQ, 11: basked smoker how things were going regarding smokingQ, 12: basked what you could do to helpQ, 18: bhelped smoker to calm downQ, 19: bsuggested nicotine replacementQ, and 20: bsuggested smoker use exerciseQ); rewarding behaviors (items 16: bcelebrated with smokerQ, 17: bpraised or encouraged smokerQ, 22: btried to minimize smoker’s stressQ, 26: bavoided conflict with smokerQ, and 27: bwere patient with smokerQ); and self-oriented behaviors (items 4: bread information about smokingQ, 8: basked another person to help you to help smokerQ, 13: bfelt proud of yourselfQ, 15: bcoped with your own stressQ, and 28: bavoided conflictQ). In addition to Cohen’s model noted above, the inclusion of self-oriented behaviors as a content domain was based on the results of studies investigating supportive behaviors designed to assist a spouse in helping their partner to seek treatment for problem drinking and other drug use (see Smith & Meyers, 2004 for review). Thomas, Yoshioka, and Ager (1993) developed unilateral family therapy to assist cooperative, non-alcohol abusing spouses to influence their alcoholic partners to stop drinking, enter treatment or both. This approach assumes that by assisting the cooperative spouse with learning coping strategies to reduce stress and anxiety about their alcoholic partner’s drinking (self-oriented behavior), the spouse’s efforts can then be channeled toward more specific behaviors to address their partner’s drinking problem. In order for the items to have generalizability to those with limited reading skills, generation of items containing three or more syllable words was avoided. Further, content validity was accomplished through regular consultation and review of the measure at all phases of development with experts in the field of social support and smoking cessation (i.e., Drs. Edward Lichtenstein and Donald Baucom).

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1355

2.1.2. Response options and scoring All positive and negative behaviors were presented within the inventory in a random order. Each item had a three-level response set, indicating whether the behavior occurred during the prior 2-week period (i.e., bYesQ, bNoQ, or bI don’t knowQ). The SPM total score was calculated by summing the number of items endorsed in the direction of supportive behaviors. This score could take on values from 0 to 29. Items that were left unanswered were considered non-endorsed. For items 3 (bnagged or preachedQ), 23 (battempted to hide cigarettesQ), 24 (bpunished or withheld rewardsQ), 25 (bcriticized or blamedQ), and 29 (bencouraged smoker to smokeQ) answering bNoQ or bI don’t knowQ was considered endorsement. For the latter response option, if subjects had performed the behavior, we reasoned they would have knowledge that they did so. Nonetheless, we recognize subjects may not know the consequences of their behavior (e.g., encouraged smoker to smoke). For the remaining items, responding bYesQ was considered endorsement. The frequency of endorsement of bI don’t knowQ ranged from 2% to 8% for the 29 items. Thus, we chose not to set as missing those items endorsed as bI don’t knowQ since our criteria for a scorable SPM was less than 20% missing items (see below). Thus including bI don’t knowQ responses as missing would result in a markedly diminished sample size for the analyses. 2.2. Evaluation of the SPM in a college young adult sample 2.2.1. Participants A total of 2057 students enrolled in one of three Midwestern undergraduate schools were invited to complete the study survey. This study was approved by the respective Institutional Review Boards at the Mayo Clinic and the three universities. Potential respondents indicated their consent to participate by completion of the survey. Their involvement was completely voluntary and surveys were completed anonymously. Because the survey was administered in the classroom, students of any age may have completed the survey. Thus, to assess the behaviors of young adults, the analysis was restricted to those students aged 18 to 24 years old. Of the 2057 respondents, 1715 responded to a question on whether or not they had a close relationship with someone whom they thought should quit smoking who did not have missing demographic data and were in the 18– 24 year age range. Of these 1715, 786 reported having a close relationship with someone whom they thought should quit smoking (i.e., concerned others). These 786 concerned others versus those who reported they did not have a close relationship with someone whom they thought should quit smoking were similar based on characteristics of age (19.9 F 1.4 versus 19.9 F 1.5 years; Z = 0.712; P = 0.476), race (89% Caucasian versus 91%; v 2(5) = 4.31; P = 0.505), marital status (96% were never married in each group; v 2(2) = 0.139; P = 0.933), year in college (35% versus 37% were in their first year of college; v 2(3) = 1.81; P = 0.612), and tobacco use (31% versus 26% were current smokers; v 2(2) = 5.25; P = 0.072). Fifty-two percent of the females versus 43% of the males reported having a close relationship with a smoker whom they thought should quit (v 2(1) = 13.4; P b 0.001). Of the 786 bconcerned otherQ subjects, 771 completed a scorable (i.e., less than 10% missing items) SPM. Thus, the 771 subjects who form the basis of this report (1) were between the ages of 18 and 24 years,

1356

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

(2) reported they had a close relationship with someone whom they thought should quit smoking, and (3) had completed a scorable SPM. 2.2.2. Procedure This study was conducted in February and March 2003 as part of a larger survey project which included an examination of the role of body image, perceived stress, and self-esteem on the smoking behaviors of young adults (Croghan et al., 2003). Classroom instructors at each of the three institutions were given a general informational form, in advance, that described the purpose of the study. If an instructor indicated interest in study participation, he or she was asked to provide course information to the investigators concerning the course title, number of students, and gender composition of the class. The survey instruments were distributed in a classroom setting by the class instructor or a study assistant. Students were recruited from a variety of college courses (i.e., psychology, biology) and level (i.e., introductory to advanced). All potential respondents were given a packet in the classroom containing the survey booklet, a cover letter explaining the purpose of the study, and an envelope to place all of the materials inside. Students were asked to complete and return the survey in class. Students who had completed the survey in a previous class were asked to refrain from completing an additional survey. The survey took approximately 30 min to complete, which was in accordance with the time frame derived from an initial pilot of the survey to determine feasibility. No identifying information was collected as part of the study, and no attempts to further contact the participants were made following return of the survey packet. 2.2.3. Measures 2.2.3.1. Demographics. Demographic characteristics assessed included age, gender, ethnicity, marital status, and year in college. 2.2.3.2. Tobacco use. All survey participants were asked about their use of cigarettes, chewing tobacco/snuff, cigars, and pipe tobacco using standard questions (CDC, 2002). For each product, participants were asked if they used the tobacco product at least 100 times in their lifetime. Participants who reported not using any tobacco products more than 100 times in their lifetime were categorized as never tobacco users. Those who reported using at least 100 times were asked to report use in the past 30 days. Of these respondents, those who reported smoking any cigarettes within the past 30 days were categorized as current cigarette smokers. Those who reported using tobacco at least 100 times, but not within the past 30 days, were categorized as former tobacco users. Because of the small number of participants in the categories of former tobacco users and current use of tobacco products other than cigarettes, respondents in these two categories were combined for the analysis as bother Q and were compared with never tobacco users and current cigarette smokers. 2.2.3.3. Concern about a cigarette smoker. Concern about a smoker was assessed by the question: bIs there someone close to you (e.g., parent, spouse, or partner) who smokes

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1357

cigarettes whom you think should quit?Q Those responding affirmatively (i.e., concerned others) were directed to complete the Support Provided Measure regarding support provided to this smoker (see description above). They were also asked to indicate their relationship to this smoker (e.g., parent, spouse/partner), the gender of the smoker, and whether they were living in the same household with the smoker. To assess reported willingness to help, respondents were also asked bWould you be interested in helping this smoker to stop smoking?Q with response options of: definitely would not want to help (0), probably not want to help (1), unsure (2), probably would want to help (3), and definitely would want to help (4). 2.2.4. Statistical methods The two-sample rank sum test and chi-square test were used to compare demographics between those subjects who knew a smoker who should quit smoking and those who did not. The proportion of survey respondents endorsing each of the 29 items of the SPM was summarized for all survey participants. Spearman rank correlation coefficients were calculated for each item with the total score (for items 3, 23, 24, 25, and 29 the reverse score (bNoQ or bI don’t knowQ) for each item). To characterize the factor structure of the SPM, principal component factor analysis was conducted on the 29 items using an orthogonal (varimax) rotation with the prior communality estimate for each variable as its squared multiple correlation (SMC) with all other variables. Based on the results of the above analyses we investigated items that could potentially be removed from the SPM based on the following criteria: (1) floor and ceiling effects (b 5% or N 95% endorsed), (2) item total score correlation (investigated items with low item total score correlation (|r| b 0.20), and (3) factor analysis results (items that did not load on any factor (|loading| b 0.30) or items that loaded on multiple factors were investigated for removal). Following removal of items, SPM scores were recalculated by summing the number of endorsed items. SPM scores were compared between select groups (e.g., smoking status, gender, type of relationship with the smoker), using the two-sample rank sum test and the Kruskal–Wallis test where appropriate. Multiple regression was used to examine the multivariate relationship between SPM score and age, gender of the respondent, relationship to the smoker, tobacco use status of the respondent and whether the respondent lived with their smoker. In all cases, statistical tests yielding p-values V 0.050 were considered as evidence of findings not attributable to chance.

3. Results 3.1. Participants As detailed in Table 1, the survey respondents (N = 771) were primarily Caucasian, single, females (89% Caucasian, 96% never married, 67% female). Their average age (FS.D.) was 19.9 F 1.4 years and their year in college was distributed relatively evenly with 35%, 32%, 19%, and 14% in years 1–4, respectively. The majority of survey respondents had no history

1358

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

Table 1 Respondent demographic characteristics (N = 771) Demographics Age, years Mean F S.D., median, range Gender Female Male Ethnicity Caucasian Asian African American/Black Native American/Alaskan Native Hispanic Other Missing Marital Status Never married Married Other Missing Year in college 1st year 2nd year 3rd year 4th year or higher Missing Tobacco history Current smoker Never user Other (former or use other types of tobacco) Lives with smokerb Yes Relationship to smokerb Boyfriend/girlfriend/spouse Parent Other family member Friend Other Gender of smokerb Female Male Missing Level of interest in helping this smokerb to quit Definitely would want to help Probably would want to help Unsure Probably would not want to help Definitely would not want to help Missing

%a, (#) 19.9 F 1.4, 20.0, 18.0–24.0 67 (513) 33 (258) 89 (686) 5 (35) 2 (12) 2 (13) 2 (13) 1 (11) 1 96 (739) 2 (13) 2 (17) 2 35 32 19 14 1

(273) (248) (145) (104)

30 (228) 62 (475) 9 (68) 29 (223) 14 (107) 41 (314) 18 (136) 20 (157) 7 (57) 45 (346) 55 (416) 9 54 (406) 27 (201) 15 (115) 3 (20) 1 (8) 21

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1359

of tobacco use (62%). Among current smokers, 38% reported daily smoking, 27% smoked on 20–29 days, 8% smoked on 10–19 days, 9% smoked on 5–9 days, and 18% smoked less than 5 of the 30 days prior to completing the survey. Further, the average number of cigarettes smoked on smoking days was less than 5 for 51%, 5–9 for 26%, 10–19 for 19%, and only 5% smoked 20 or more cigarettes per day. Of the smokers that the respondents thought should quit, 55% were reported by the respondent to be male; 41% were a parent, 20% a friend, 18% another family member, and 14% a romantic partner or spouse; and 29% lived with the respondent. Overall, 54% of the respondents indicated they bdefinitelyQ would want to help their smoker quit smoking. 3.2. Item endorsement Table 2 describes the proportion of participants endorsing each of the individual items of the SPM. Item response rates ranged from 6% to 95% for the 29 items. Only 6% celebrated with their smoker for their efforts to quit smoking or to stay smoke-free and 11% felt proud of themselves for helping their smoker to quit or to stay smoke-free in the past 2 weeks. In contrast, 95% of respondents indicated that they did not encourage their smoker to smoke, 83% did not attempt to hide or keep cigarettes away from their smoker; and 83% did not criticize or blame their smoker for consequences related to their smoking. Further, 44% of the participants expressed their concern about their smoker’s health or the health effects of smoking on others, 43% discouraged their smoker from smoking in certain situations/places, 40% coped with their stress by relaxing, taking a break, taking a walk, or calling a friend, and 40% were patient with the emotional ups and downs of their smoker. All but 5 of the 29 items were positively correlated with the total SPM score. Spearman correlation coefficients ranged from 0.37 to 0.65 (Table 2). The following items were negatively correlated with the total score: items 3: bNagged or preached to your smoker about his/her smoking, quitting, or staying smoke-free?Q (r = 0.37), 23: bAttempted to hide or keep cigarettes away from your smoker?Q (r = 0.26), 24: bPunished your smoker or withheld rewards (e.g., affection) for reasons related to his/her smoking or staying smoke-free?Q (r = 0.25), 25: bCriticized or blamed your smoker for the consequences of his/her smoking?Q (r = 0.24), and 29: bEncouraged your smoker to smoke?Q (r = 0.06). The positively correlated items ranged from r = 0.33 for item 16: bCelebrated with your smoker for his/her efforts to quit smoking or stay smoke-free?Q to r = 0.65 for item 21: bExpressed concern about your smoker’s health or the health effects of smoking on others?Q Notes to Table 1: a For each variable, the data presented correspond to the number (%) within each response category unless indicated otherwise. Percentage may not total to 100 due to rounding. b Smoker refers to someone close to the respondent who smokes cigarettes whom the respondent thinks should quit.

1360

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

Table 2 SPM item endorsement rates and total score correlations ra

Item no.

% Endorsed

(1) Asked your smoker if he/she was willing to discuss their smoking or staying smoke-free? (2) Discussed quitting or staying smoke-free with your smoker? (3) Nagged or preached to your smoker about his/her smoking, quitting, or staying smoke-free?b (4) Read information on smoking, quitting, or staying smoke-free? (5) Gave your smoker information on effects of smoking, quitting, or staying smoke-free? (6) Tried to model healthy behavior for your smoker by exercising? (7) Tried to model healthy behavior for your smoker by following a healthy diet? (8) Asked another person to help your smoker quit or stay smoke-free? (9) Helped your smoker to come up with ideas to help him/her quit smoking or to stay smoke-free? (10) Discouraged your smoker from smoking in certain situations/places (e.g., by making your home smoke-free)? (11) Asked your smoker how things were going regarding his/her smoking, quitting, or staying smoker-free? (12) Asked your smoker what you could do to help him/her quit smoking or stay smoke-free? (13) Felt proud of yourself for helping your smoker quit smoking or stay smoke-free? (14) Engaged in a smoke-free activity with your smoker (e.g., exercised, went to a movie, ate at a smoke-free restaurant)? (15) Coped with your stress by relaxing, taking a break, taking a walk, or calling a friend? (16) Celebrated with your smoker for his/her efforts to quit smoking or stay smoke-free? (17) Praised or encouraged your smoker for his/her efforts to quit smoking or stay smoke-free? (18) Helped your smoker to calm down when he/she was feeling stressed? (19) Suggested that your smoker use medication to become or to stay smoke-free (e.g., nicotine patch, nicotine gum)? (20) Suggested that your smoker use exercise to become or to stay smoke-free? (21) Expressed your concern about your smoker’s health or the health effects of smoking on others? (22) Tried to minimize your smoker’s stress level (e.g., by taking over household responsibilities) to help him/her to quit smoking or to stay smoke-free? (23) Attempted to hide or keep cigarettes away from your smoker?b (24) Punished your smoker or withheld rewards (e.g., affection) for reasons related to his/her smoking or staying smoke-free?b (25) Criticized or blamed your smoker for the consequences of his/her smoking?b (26) Avoided conflict with your smoker in an effort to help him/her quit smoking or stay smoke-free? (27) Been patient with the emotional ups and downs of your smoker? (28) Spent time thinking about your smoker’s smoking, quitting, or staying smoke-free? (29) Encouraged your smoker to smoke?b

28

0.57

36 63 (37)

0.60 0.37

17 13

0.34 0.39

32 26 17 19

0.50 0.48 0.41 0.55

43

0.56

27

0.59

21

0.61

11 38

0.40 0.55

40

0.57

6

0.33

24

0.59

33 14

0.56 0.39

21 44

0.49 0.65

17

0.49

83 (17) 93 (7)

0.26 0.25

82 (18) 17

0.24 0.44

40 39

0.60 0.59

95 (5)

0.06

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1361

3.3. Factor analysis A principal component factor analysis was performed on all 29 of the SPM items. The principal component factor analysis with a varimax rotation (Table 3) yielded a two-factor solution accounting for 84% of the variance. Only 2 of the 29 items did not load on either factor, item 4: bRead information on smoking, quitting, or staying smoke-free?Q and item 29: bEncouraged your smoker to smoke?Q Factor 1 (eigenvalue = 7.5), comprising 20 items, was interpreted as bSmokerQ items and accounted for 73% of the variance. Of the 20 items on factor 1, 7 had loadings less than 0.40. These items had loadings from 0.34 to 0.39 in absolute value. Factor 2 (eigenvalue = 1.2), consisting of 7 items, was interpreted as bSelfQ items and accounted for 11% of the variance. All but 1 of the 7 items (item 26: avoided conflict with smoker, item loading = 0.35) had loadings z0.40. Items 18, 20, 22, 26, and 28 loaded on each factor. 3.4. Item reduction Our item reduction was based on the following criteria: (1) floor and ceiling effects (b 5% or N 95% endorsed), (2) item total score correlation (we investigated items with item total score |correlation| b 0.20, and/or (3) factor analysis results (items that did not load on any factor (|loading| b 0.30) or items that loaded on multiple factors were investigated). Based on these criteria the following 7 items were eliminated for the following reasons: item 4: did not load on either factor; items 18, 20, 22, 26, and 28: loaded on both factors with approximately equal loadings; and item 29: did not load on either factor and had a very high endorsement rate (95%). Following the item reduction, the SPM total score could take on values from 0 to 22. The SPM total score was 8.0 F 3.8 (median = 7.0; range = 2–19). Factor 1, smoker items, had a total score of 6.2 F 2.8 (median = 5.0; range = 2–16) and factor 2, self-items, had a total score of 1.8 F 1.6 (median = 1.0; range = 0–5). 3.5. Internal consistency reliability Reliability was assessed by calculating Kuder–Richardson formula 20 (KR20). Following elimination of the 7 items above, the KR20 coefficient for the remaining 22 items was 0.77. The KR20 for factors 1 and 2 was 0.68 and 0.71, respectively. Notes to Table 2: Missing values ranged from 0 to 3 participants for each item. Items left blank are considered non-endorsement. For all items yes is endorsement except for items 2, 23, 24, 25, and 29. For these items, the % who responded byesQ appears in ( ). Sum of endorsed items = total score. a Spearman correlation coefficient with the SPM total score (items 3, 23, 24, 25, and 29 where endorsement is bNoQ or bI don’t knowQ) was used in calculating the correlation coefficient. All correlation coefficients were significantly different from 0 ( p b 0.001 in each case) except for item 29 ( p = 0.106). b Item that was reversed scored to reflect endorsement (i.e., 95% indicated that they did not encourage their smoker to smoke and 83% did not attempt to hide or keep cigarettes away from their smoker).

1362

Table 3 Factor analysis of the support provide measure (SPM) in 771 college students

Factor 1

Factor 2

Factor 1: Smoker items (12) Asked your smoker what you could do to help him/her quit smoking or stay smoke-free? (9) Helped your smoker to come up with ideas to help him/her quit smoking or to stay smoke-free? (11) Asked your smoker how things were going regarding his/her smoking, quitting, or staying smoker free? (2) Discussed quitting or staying smoke-free with your smoker? (1) Asked your smoker if he/she was willing to discuss their smoking or staying smoke-free? (21) Expressed your concern about your smoker's health or the health effects of smoking on others? (3) Nagged or preached to your smoker about his/her smoking, quitting, or staying smoke-free? (17) Praised or encouraged your smoker for his/her efforts to quit smoking or stay smoke-free? (10) Discouraged your smoker from smoking in certain situations/places (e.g. by making your home smoke-free)? (13) Felt proud of yourself for helping your smoker quit smoking or stay smoke-free? (25) Criticized or blamed your smoker for the consequences of his/her smoking? (23) Attempted to hide or keep cigarettes away from your smoker? (19) Suggested that your smoker use medication to become or to stay smoke-free (e.g. nicotine patch, nicotine gum)? (22) Tried to minimize your smoker's stress level (e.g. by taking over household responsibilities) to help him/her to quit smoking or to stay smoke-free? (24) Punished your smoker or withheld rewards (e.g. affection) for reasons related to his/her smoking or staying smoke-free? (5) Gave your smoker information on effects of smoking, quitting, or staying smoke-free? (16) Celebrated with your smoker for his/here fforts to quit smoking or stay smoke-free? (18) Helped your smoker to calm down when he/she was feeling stressed? (8) Asked another person to help your smoker quit or stay smoke-free? (20) Suggested that your smoker use exercise to become or to stay smoke-free?

70 68 64 63 62 57 -56 54 47 44 -42 -41 41 39 -38 38 38 37 36 34

22 15 21 17 12 37 -14 30 31 24 -15 -15 14 34 -16 22 17 36 22 37

Factor 2: Self-items 2 70 (6) Tried to model healthy behavior for your smoker by exercising? 2 70 (7) Tried to model healthy behavior for your smoker by following a healthy diet? 24 48 (15) Coped with your stress by relaxing, taking a break, taking a walk, or calling a friend? 32 45 (27) Been patient with the emotional ups and downs of your smoker? 27 43 (14) Engaged in a smoke-free activity with your smoker (e.g. exercised, went to a movie, ate at as moke-free restaurant)? 40 40 (28) Spent time thinking about your smoker's smoking, quitting, or staying smoke-free? 34 35 (26) Avoided conflict with your smoker in an effort to help him/her quit smoking or stay smoke-free? -18 25 (29) Encouraged your smoker to smoke? 21 -5 (4) Read information on smoking, quitting, or staying smoke-free? Entries are item loadings multiplied by 100 and rounded to the nearest integer. For an item to be considered on a factor the loading score had to be z30 in absolute value. Items are ordered based on descending absolute value of their loading scores under the factor that they were assigned. Shading indicates items that loaded on multiple factors (i.e., bcrossloadingsQ). Items with crossloadings were placed on a given factor based on theoretical rationale.

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

Item no.

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1363

3.6. Association of SPM total and factor scores with select demographic characteristics Table 4 presents SPM total scores and factor scores by select demographics. Female participants had significantly higher SPM total scores than males ( P b 0.001) and participants that lived with their smoker had higher SPM total scores than those that did not live with their smoker ( P = 0.040). Further, those who reported their smoker was a spouse or romantic partner had higher total scores than those concerned about a parent, other family member, friend or other person ( P b 0.001). However, no significant univariate differences were identified when total scores on the SPM were compared across smoking status ( P = 0.946) nor was there an age association ( P = 0.424). Similar results were obtained for each factor. From a multivariate analysis including respondent age, gender, and tobacco use status; whether or not the respondent lived with their smoker, type of relationship to their smoker, and each two-way interaction, a significant gender by relationship interaction was found ( F = 11.13; P b 0.001). Following this, the model with each main effect (age, gender, whether or not the respondent lived with their smoker, type of relationship to the smoker, and tobacco use status) and the gender by relationship interaction was fit. From this model, respondents that lived with their smoker ( F = 6.0; P = 0.015) and those who were romantically involved ( F = 47.71; P b 0.001) reported higher SPM total scores. The gender by relationship interaction was also significant ( F = 10.4; P = 0.001) with male respondents, indicating higher SPM total scores when their relationship with their smoker was romantic or marital.

4. Discussion This study indicates that the Support Provided Measure (SPM), when completed by young adult college students ages 18–24, has good internal consistency reliability (KR20 = 0.77). Further, the measure appears to assess a wide range of individual differences in the provision of supportive behaviors, as evidenced by a range of item endorsement (i.e., from 6% to 93% for the 22 retained items and total scores (2 to 19)). The scale in this sample was best characterized with two factors. It is interesting that the first factor emphasized behaviors done primarily to help the smoker, while the second factor emphasized behaviors the support person could do for himself or herself. This may indicate that different functions of support are associated with different behaviors and could be targets of future interventions. Results of the multivariate analysis indicate that young adult males reported providing more support to their smoker when they were concerned about a romantic partner or spouse. Prior studies indicate females are more willing to assist other family members with health behavioral change (Umberson, 1992) but our findings indicate that younger males might be engaged in supportive interventions when the focus is a spouse or partner who smokes. Our findings indicate no effect of the support person’s smoking status on the amount of reported support provided. Some (e.g., Britt, Curry, McBride, Grothaus, & Louie, 1994) but not all (e.g., McBride et al., 1999) previous studies indicate nonsmokers are more effective as supporters with respect to the smoker achieving abstinence. In the Lung Health Study (Murray et al., 1995), subjects supported by a smoker were less than half as likely to stop

1364

Table 4 SPM total and factor scores by select demographics StatisticT 6.1

P

8.5 F 3.8; 8.0 7.0 F 3.5; 6.0 2.1

4.0

0.040

P

78.4 (4)

0.112 (2)

0.077

b 0.001

8.0 4.0 4.0 6.0 5.0

0.946

7.9 F 3.6; 7.0 8.2 F 4.3; 7.0 8.1 F 3.6; 8.0

1.02 (2)

0.424

2.3

0.021

b 0.001

2.0 1.0 1.0 2.0 2.0

0.600

1.27 (2)

0.529

1.35 (2)

0.509

1.8 F 1.6; 2.0 1.7 F 1.6; 1.0 1.7 F 1.7; 1.0 3.43 (2)

6.0 F 2.6; 5.0 6.4 F 2.9; 5.0 6.3 F 2.8; 5.0

b 0.001

52.5 (4) 2.4 F 1.6; 1.4 F 1.5; 1.6 F 1.5; 2.2 F 1.6; 1.9 F 1.7;

6.1 F 2.6; 5.0 6.5 F 3.1; 5.0 6.3 F 2.5; 6.0 1.72 (2)

P

7.0

1.7 F 1.6; 1.0 2.0 F 1.7; 2.0 71.2 (4)

7.9 F 3.0; 5.7 F 2.5; 5.4 F 2.1; 6.8 F 2.9; 6.4 F 2.9;

StatisticT

2.0 F 1.6; 2.0 1.2 F 1.5; 1.0 1.8

b 0.001

Factor 2

b 0.001

6.1 F 2.7; 5.0 6.5 F 2.9; 5.0

10.0 6.0 6.0 8.0 7.0

7.8 F 3.7; 7.0 8.2 F 3.9; 7.0 7.8 F 3.8; 7.0

StatisticT

6.5 F 2.8; 5.0 5.8 F 2.6; 4.0

7.8 F 3.7; 7.0 8.5 F 4.1; 8.0 10.2 F 3.9; 7.1 F 3.5; 6.9 F 3.2; 9.0 F 3.9; 8.3 F 4.1;

Factor 1

b 0.001

0.180 1.8 F 1.6; 2.0 1.8 F 1.6; 1.0 1.5 F 1.5; 1.0

T Analysis was performed using the two-sample rank sum test or Kruskal-Wallis as appropriate. Statistic presented is the Z statistic from two-sample rank sum test or Chi-square statistic with degrees of freedom in paranthesis for Kruskal-Wallis test.

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

Total score Gender Females Males Does the smoker live with you? No Yes Relationship to smoker Boyfriend/girlfriend/spouse Parent Other family member Friend Other Smoking status Never Current Former/other tobacco use Age 18–20 21–22 23–24

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1365

smoking than those supported by a nonsmoker. However, less attention has focused on younger individuals as support persons. Thus, studies are needed to test whether younger individuals can provided effective support, defined as assisting their smoker to quit, irrespective of their personal smoking status. The potential role of young adults in assisting a parent, friend, or other family member in making a quit attempt is indicated by the observation that over half of the sample was bdefinitelyQ interested in helping a smoker to quit. Further, of the smokers the young adults were interested in helping, over half had a familial relationship with their smoker and the remaining indicated a friend, spouse, or romantic partner they wanted to help. Interventions delivered using young adults as change agents might be a novel public health approach to intervene with smokers of all ages, including young adults. Several methodological limitations of the present investigation should be considered when interpreting the results and that need to be addressed in future studies. First, we relied on a convenience sample of primarily Caucasian, young adult college students, limiting generalizability to other populations. All respondents completing the SPM indicated whether they bwere close to someone who smoked cigarettes whom they thought should quitQ (yes/ no). The complex wording and compound nature of this dichotomous question limits our ability to know if the responses relate to someone whom they are both close to and for whom they think should quit smoking. Future studies could include a design whereby SPM scores are compared for individuals who know a smoker but differ based on whether or not they have a close relationship to the smoker. Second, we recommend that future studies eliminate the bI don’t knowQ response option, restricting the response set to a dichotomous (yes/no) and/or a continuous likert-type response scale, indicating frequency of the behaviors that have occurred. Third, a more stringent validation of the SPM should include an assessment of test–retest reliability and concurrent or criterion group validity. The design and sample limitations of this study did not permit such evaluation. While we considered restricting the sample to those who were not only concerned about a smoker and reported they were willing to help this smoker quit, the intent of this study was to examine the psychometric properties of this instrument in a broader, more generalizable sample. In addition, limiting the sample to those willing to help their smoker quit would have likely restricted the range of SPM scores to the higher end. Although first tested among young adults, our long-term goal is to develop a measure that will be applicable to a wide range of clinic-based and population-based samples. We recommend that future studies examine the psychometric properties of the SPM in samples of support persons or concerned others who are willing to help their smoker quit. Further, although the items were selected through empirically based content validation methods, there are undoubtedly important supportive and nonsupportive behaviors that were not included in the measure. Thus, further attempts to describe the diversity of supportive and non-supportive behaviors for smoking cessation are encouraged. Moreover, achieving an approximately equal number of items for each content domain would overcome the limitations of the current study. Future scale development could also be enhanced through analysis of the predictive validity of the individual items and total score in association with smoking abstinence and quit attempts. This empirical process would improve the content validity of the item pool and may allow for a shorter version of the

1366

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

measure. Fourth, the items endorsed on the SPM may be susceptible to social desirability response bias. A fifth limitation of the study is that while we assessed concerned other reports of support provided to a smoker, we did not collect data from their smokers to determine readiness level to quit smoking. Future studies could assess whether SPM scores are associated with the smoker’s motivation to quit. Related to this, we did not assess the perception of support received by the smokers in the sample. Given the design and intention of the current study, this was not feasible. Ultimately, it will be important to assess both the perception of the provision and the receipt of support. The support efforts of concerned others may be ineffective, misinterpreted, or unrecognized by the smoker. It is important to note that some of the SPM items reflect activities of the support person that may never be known by the smoker. Seven of the retained items (i.e., 6, 7, 8, 13, 15, 23, 27) reflect activities the support person could do entirely independent of their smoker. Research is needed to clarify the intended or actual, as opposed to the perceived, supportive behaviors provided to and received by a smoker. Although some studies have compared these perceptions (Lichtenstein et al., 2002; Pollak et al., 2001), few smoking specific supportive behaviors were assessed. Knowing how perceived support behaviors differ between the support provider and the support receiver could help in designing interventions to account for these gaps. Attempts have been made to develop observational coding schemes to assess detailed information regarding the behaviors spouses engage in when attempting to provide support to their marital partners (Barbee, 1990; Cutrona, Suhr, & MacFarlane, 1990; Liotta, Jason, Robinson, & LaVigne, 1985). Thus, the analyses of supportive transactions could be utilized to validate the SPM against objective observable behaviors. To obtain more objective data, real time evaluation of support transactions may assist in identifying differential support elicitation techniques used. For example, a smoker and their support person could be instructed to btalk about smoking or quitting smoking for 20 min.Q During this interaction, observers trained in behavioral observation could code all verbal and non-verbal transactions. 4.1. Future directions With funding from the National Institute on Drug Abuse, efforts are currently under way to further refine the SPM. Given that the goal of this study was to develop and evaluate a measure of support provided for smoking cessation, these findings will serve as preliminary data to further develop the measure. Our ongoing work includes the development of a companion measure of support received for smoking cessation. This will enable us to examine the association of these two measures of supportive behaviors or appropriate subsets of items, with smoking cessation outcomes among smokers in other study populations. Moreover, in future work we plan to assess the association between SPM scores and smoker readiness to stop smoking. Further, the new measures could be utilized in clinical and public health intervention trials designed to modify behaviors among concerned others in the smoker’s natural environment to enhance abstinence rates in smokers. Individual differences in the support person associated with supportive behaviors provided (e.g., depression, marital satisfaction) should be examined as possible contributing variables to support provided. For

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1367

example, distressed and nondistressed marital couples may differ in the types and level of the support they provide and report receiving (Palmer et al., 2000). Finally, future studies could address the level of relationship satisfaction associated with the reported provision and receipt of support during smoking cessation. Thus, generalizability of the current findings, including the factor structure and other scale properties, must be assessed in larger, more diverse population-based and clinical samples.

Acknowledgements The authors wish to acknowledge the contributions of Drs. Don Baucom and Ed Lichtenstein for their assistance with the content validation of the Support Provided Measure.

References Albrecht, S., Payne, L., Stone, C. A., & Reynolds, M. D. (1998). A preliminary study of the use of peer support in smoking cessation programs for pregnant adolescents. Journal of the American Academy of Nurse Practitioners, 10(3), 119 – 125. Barbee, A. P. (1990). Interactive coping: The cheering up process in close relationships. In S. Duck, & R. C. Silver (Eds.), Personal Relationships and Social Support (pp. 46 – 65). London7 Sage. Black, D. R., Tobler, N. S., & Sciacca, J. P. (1998). Peer helping/involvement: An efficacious way to meet the challenge of reducing alcohol, tobacco, and other drug use among youth. Journal of School Health, 68(3), 87 – 93. Boyle, R. G., O’Connor, P., Pronk, N., & Tan, A. (2000). Health behaviors of smokers, ex-smokers, and never smokers in an HMO. Preventive Medicine, 31, 177 – 182. Breteler, M. H. M., Schotborg, E. J., & Schippers, G. M. (1996). The effectiveness of smoking cessation programs: Determinants and outcomes. Psychology & Health, 11(1), 133 – 153. Britt, J., Curry, S. J., McBride, C., Grothaus, L., & Louie, D. (1994). Implementation and acceptance of outreach telephone counseling for smoking cessation with nonvolunteer smokers. Health Education Quarterly, 55 – 68. Carlson, L. E., Goodey, E., Bennett, M. H., Taenzer, P., & Koopmans, J. (2002). The addition of social support to a community-based large-group behavioral smoking cessation intervention: Improved cessation rates and gender differences. Addictive Behaviors, 27(4), 547 – 559. Centers for Disease Control and Prevention. (2002). Cigarette smoking among adults—United States, 2000. Morbidity and Mortality Weekly Report, 51(29), 642 – 645. Cohen, S., & Lichtenstein, E. (1990). Partner behaviors that support quitting smoking. Journal of Consulting and Clinical Psychology, 58(3), 304 – 309. Collins, R., Emont, S. L., & Zywiak, W. H. (1990). Social influence processes in smoking cessation: Postquitting predictors of long-term outcome. Journal of Substance Abuse, 2(4), 389 – 403. Croghan, I. T., Patten, C., Bronars, C., Schroeder, D., Nirelli, L., Foraker, R., et al. (2003). Relationship of smoking and gender with body image, perceived stress, and self-esteem in young adults. 5th Annual European Conference of the Society for Research on Nicotine and Tobacco. Padua, Italy, November 20–22. Cutrona, C. E., Suhr, J. A., & MacFarlane, R. (1990). Interpersonal transactions and the psychological sense of support. In S. Duck, & R. Silver (Eds.), Personal Relationships and Social Support (pp. 30 – 45). London7 Sage Publications. Elwood, R. W., & Jacobson, N. S. (1988). The effects of observational training on spouse agreement about events in their relationship. Behavior Research and Therapy, 26(2), 159 – 167.

1368

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

Fiore, M. C., Bailey, W. C., Cohen, S. J., Dorfman, S. F., Goldstein, M. G., Gritz, E. R., et al. (2000). Treating Tobacco Use and Dependence. Clinical Practice Guideline. Rockville, MD7 US Department of Health and Human Services, Public Health Service. Fisher Jr., E. B. (1997). Two approaches to social support in smoking cessation: Commodity model and nondirective support. Addictive Behaviors, 22(6), 819 – 833. Gruder, C. L., Mermelstein, R. J., Kirkendol, S., & Hedeker, D. (1993). Effects of social support and relapse prevention training as adjuncts to a televised smoking-cessation intervention. Journal of Consulting and Clinical Psychology, 61(1), 113 – 120. Hanson, B. S., Isacsson, S. O., Janzon, L., & Lindell, S. E. (1990). Social support and quitting smoking for good. Is there an association? Results from the population study, bMen born in 1914,Q Malmo, Sweden. Addictive Behaviors, 15(3), 221 – 233. Herzog, T. A., Abrams, D. B., Emmons, K. M., & Linnan, L. (2000). Predicting increases in readiness to quit smoking: A prospective analysis using the contemplation ladder. Psychological Health, 15, 369 – 381. Lakey, B., & Cohen, S. (2000). Social support theory and measurement. In S. Cohen, L. Underwood, & B. Gottlieb (Eds.), Social Support Measurement and Intervention. New York7 Oxford University Press. Lichtenstein, E., Andrews, J. A., Barckley, M., Akers, L., & Severson, H. H. (2002). Women helping chewers: Partner support and smokeless tobacco cessation. Health Psychology, 21, 273 – 278. Lichtenstein, E., Glasgow, R., & Abrams, D. (1986). Social support in smoking cessation: In search of effective interventions. Behavior Therapy, 17(5), 607 – 619. Lichtenstein, E., & Hollis, J. F. (1992). Patient referral to smoking cessation programs: Who follows through? Journal of Family Practice, 34, 739 – 744. Liotta, R. F., Jason, L. A., Robinson, W., & LaVigne, V. (1985). A behavioral approach for measuring social support. Family Therapy, 12(3), 285 – 295. McBride, C. M., Curry, S. J., Lando, H. A., Pirie, P. L., Grothaus, L. C., & Nelson, J. C. (1999). Partner smoking status and pregnant smoker’s perceptions of support for and likelihood of smoking cessation. Health Psychology, 17, 63 – 69. Mermelstein, R., Lichtenstein, E., & McIntyre, K. (1983). Partner support and relapse in smoking cessation programs. Journal of Consulting and Clinical Psychology, 51, 456 – 466. Murray, R. P., Johnston, J. J., Dolce, J. J., Lee, J. J., & O’Hara, P. (1995). Social support for smoking cessation and abstinence: The Lung Health Study. Addictive Behaviors, 20(2), 159 – 170. Palmer, C. A., Baucom, D. H., & McBride, C. M. (2000). Couple approaches to smoking cessation. In T. Schmaling, & T. G. Sher (Eds.), The Psychology of Couples and Illness (pp. 311 – 336). Washington, DC7 American Psychological Association. Park, E. W., Tudiver, F., Schultz, J. K., & Cmpbell, T. (2004). Does enhancing partner suport and interaction improve smoking cessation? A meta-analysis. Annals of Family Medicine, 2, 170 – 174. Patten, C. A., Offord, K. P., Hurt, R. D., Cox, L. S., Thomas, J. L., Quigg, S. M., et al. (2004). Training support persons to help smokers quit: A pilot study. American Journal of Preventive Medicine, 26(5), 386 – 390. Patten, C. P., Lopez, K. N., Thomas, J. L., Offord, K. P., & Decker, P. A. (2004). Reported willingness among adolescent nonsmokers to help parents, peers, and others to stop smoking. Preventive Medicine, 39, 1099 – 1106. Pirie, P. L., Rooney, B. L., Pechacek, T. F., Lando, H. A., & Schmid, L. A. (1997). Incorporating social support into a community-wide smoking cessation contest. Addictive Behaviors, 22(1), 131 – 137. Pollak, K. I., McBride, C. M., Baucom, D. H., Curry, S. J., Lando, H., Pirie, P. L., et al. (2001). Women’s perceived and partners’ reported support for smoking cessation during pregnancy. Annals of Behavioral Medicine, 23(3), 208 – 214. Prince, F. (1995). The relative effectiveness of a peer-led and adult-led smoking intervention program. Adolescence, 30(117), 187 – 194. Roski, J., Schmid, L. A., & Lando, H. A. (1996). Long-term associations of helpful and harmful spousal behaviors with smoking cessation. Addictive Behaviors, 21(2), 173 – 185. Smith J. E., & Meyers R. J. (Eds.) (2004). Motivating Substance Users to Enter Treatment: Working with Family Members. New York7 Guilford Press.

J.L. Thomas et al. / Addictive Behaviors 30 (2005) 1351–1369

1369

Thomas, E. J., Yoshioka, M., & Ager, R. D. (1993). The spouse sobriety influence inventory. In K. Corcoran, & J. Fischer (Eds.), Measures for Clinical Practice: A Sourcebook (2nd ed.). New York7 Free Press. Thomas, J. L., Patten, C. A., Offord, K. P., & Decker, P. A. (2004). Preliminary finding on the development of a measure of supportive behaviors provided by support persons to help someone stop smoking. Journal of Behavioral Medicine, 27(5), 507 – 523. Umberson, D. (1992). Family status and health behaviors: Social control as a dimension of social integration. Journal of Health and Social Behavior, 28, 306 – 319.