Need for Cognition as a Predictor and a Moderator of Outcome in a Tailored Letters Smoking Cessation Intervention

Transcription

1 Health Psychology 2010 American Psychological Association 2010, Vol. 29, No. 4, /10/$12.00 DOI: /a Need for Cognition as a Predictor and a Moderator of Outcome in a Tailored Letters Smoking Cessation Intervention Severin Haug, Christian Meyer, Sabina Ulbricht, and Beatrice Gross Ernst-Moritz-Arndt-University Greifswald Ulrich John Ernst-Moritz-Arndt-University Greifswald Hans-Jürgen Rumpf University of Lübeck Objective: To analyze whether baseline need for cognition (NFC) was a predictor or a moderator of treatment outcome in a tailored letters intervention for smoking cessation. Design: A total of 1,499 daily smokers were recruited from general medical practices in Germany within a quasi-randomized trial testing the efficacies of two brief interventions for smoking cessation: (a) computer-generated tailored letters and (b) physician-delivered brief counseling versus assessment-only. For this study, we used data from 1,097 daily smokers who were assigned to the tailored letters or the assessment-only condition. Main Outcome Measures: self-reported 6-month prolonged abstinence from tobacco smoking assessed at 12-, 18-, and 24-month follow-ups, and smoking cessation self-efficacy assessed at 6- and 24-month follow-ups. Results: Baseline NFC predicted 6-month prolonged smoking abstinence ( p.01) and smoking cessation self-efficacy ( p.01). When compared to assessment only, NFC did not moderate the effect of the tailored letters intervention on smoking abstinence ( p.05) but on smoking cessation self-efficacy ( p.05). Tailored letters resulted in higher smoking cessation self-efficacy only for persons with higher NFC. Conclusion: Higher levels of NFC are required to increase smoking cessation self-efficacy in computer-tailored interventions for smoking cessation. Considering an individual s NFC might improve the efficacy of written interventions for smoking cessation. Keywords: need for cognition, smoking cessation, brief intervention, predictors, moderators Within the framework of the elaboration likelihood model (Petty & Cacioppo, 1986), the degree to which individuals are motivated and able to process a persuasive message determines the carefulness with which the central merits of a message will be considered and evaluated. Need for cognition (NFC) refers to an individual s tendency to enjoy and engage in effortful cognitive activity (Cacioppo & Petty, 1982). Individuals with high NFC are motivated to seek information actively and to think about arguments presented to them. Individuals with low NFC pay more attention to the source of the arguments (e.g., celebrities, credible authorities, experts), the ease with which they can be processed (e.g., presented pictorially vs. verbally), and the number of arguments presented, to process information (Cacioppo, Petty, Feinstein, & Jarvis, 1996). Considering individuals NFCs in the context of health behavior interventions could be crucial for improving their outcomes. Message tailoring involves increasing the relevance of a message by Severin Haug, Christian Meyer, Sabina Ulbricht, Beatrice Gross, and Ulrich John, Institute of Epidemiology and Social Medicine, Ernst-Moritz- Arndt-University Greifswald; Hans-Jürgen Rumpf, Department of Psychiatry and Psychotherapy, University of Lübeck. Correspondence concerning this article should be addressed to Severin Haug, Ernst-Moritz-Arndt-University Greifswald, Institute of Epidemiology and Social Medicine, Walther-Rathenau-Str. 48, D Greifswald, Germany. severin.haug@uni-greifswald.de customizing it to the informational needs, interests, and concerns of a recipient. Computer-generated tailored self-help materials have been shown to be more effective to promote health behavior change than untailored materials or no intervention at all (Lancaster & Stead, 2005; Noar, Benac, & Harris, 2007). Although the exact mechanism responsible for this tailoring effect is not known, it is generally thought that compared to untailored messages, tailored messages are more likely to be attended to (Kreuter, Bull, Clark, & Oswald, 1999). Beyond tailoring to sociodemographic variables, messages have been most frequently tailored based on theoretical mechanisms associated with behavior change (e.g., self-efficacy, processes of change and stage of change) according to the Transtheoretical Model of behavior change (TTM) (Prochaska & Velicer, 1997). To date, few studies have examined the effect of tailoring messages to the recipients personality characteristics such as NFC. Two previous studies have examined the influence of NFC in the context of smoking cessation. Steward et al. (2003) reported that higher baseline NFC predicted self-efficacy to quit smoking after reading persuasive messages. However, baseline NFC was negatively associated with intention to quit, that is, low NFC individuals showed higher intention to quit after reading the messages than those with high NFC. In a separate study among college students, Vidrine et al. (2007) tested whether NFC moderated risk perception in response to a fact-based versus emotion-based smoking risk pamphlet. NFC interacted with type of message to moderate risk perceptions. However, this was the case only for occa- 367

2 368 HAUG ET AL. sional smokers. The fact-based message produced the greatest increase in risk perception among participants with higher NFC, whereas among participants with lower NFC, the emotion-based message produced the greatest change. The existing studies suggest that taking NFC into account during smoking cessation interventions might improve their outcome. However, these studies only tested the immediate effects of persuasive messages on variables related to smoking cessation (risk perception, intention to quit). They did not test longer-term effects nor did they examine the impact of NFC on smoking cessation itself. Previous studies testing predictors of smoking cessation primarily considered variables associated with nicotine dependence and those derived from health behavior change theories. They identified number of cigarettes smoked per day, duration of past quit attempts, self-efficacy, stage of change, and school education as the most important predictors (Abrams, Herzog, Emmons, & Linnan, 2000; Farkas et al., 1996; Hymowitz et al., 1997; Velicer, Redding, Sun, & Prochaska, 2007). Stable cognitive variables that may influence the processing of health messages, such as NFC, were not considered in any of these studies. We tested whether NFC could be a predictor of smoking cessation when controlling for the aforementioned established predictors. Furthermore, we also tested the moderating effect of NFC in a written tailored letters smoking cessation intervention when compared to an assessment only group. NFC acting as a moderator would suggest a different impact on smoking cessation depending on the study condition. Based on previous findings showing that a high level of thinking about an issue leads to stronger attitude change (Petty, Haugtvedt, & Smith, 1995), we expected that baseline NFC would be positively associated with smoking abstinence and smoking cessation self-efficacy at the follow-up assessments. Furthermore, we expected NFC to moderate the effectiveness of the tailored letters intervention when compared to assessment only. Persons with high NFC might be more motivated to actively think and reflect on arguments presented in the tailored letters, whereas persons with low NFC might be less motivated to seek information in the tailored letters actively and might benefit less from this intervention when compared to the assessment only condition. Method Sample Data for this study were gathered from the project Proactive interventions for smoking cessation in general medical practice (ProGP) (Meyer et al., 2008). It tested the efficacy of two brief smoking-cessation interventions against an assessment-only condition in 34 randomly selected general practices in a defined region of Northern Germany. For a period of 3 weeks all consecutive patients were screened for smoking status by a research nurse covering complete office hours. In total, we registered 11,560 practice attendances. Patients visiting the practice repeatedly within the study period (n 1664) were subsequently excluded from rerecruitment, leaving 9896 patients. From these, 36 (0.4%) patients refused to be screened by the study nurse and 302 patients were excluded for other reasons (e.g., too ill, cognitively impaired, insufficient language capabilities, screening missed). Furthermore, we excluded 7696 patients who did not fulfill the inclusion criteria, that is, age below 18 or above 70 years or not smoking cigarettes daily in the past 4 weeks. Among patients fulfilling the inclusion criteria (n 1862), 1,499 (80.5%) consented to take part. Patients were assigned to one of three study groups using a quasi-randomized procedure based on the time of attendance at the general practice. Patients in the first study week were allocated to the assessment-only control condition; patients in the second week were assigned to the tailoredletters intervention; and patients in the third week were assigned to a brief physicians advice intervention, which also included selfhelp manuals for smoking cessation. A fixed sequence of study conditions was chosen, to avoid counseling activities of the practitioner in the first and second week. The ProGP study found statistically significant effects of both interventions compared to assessment only. Among participants completing the 24-months follow-up, 6-month prolonged abstinence was 18.3% in the tailored letters intervention group, 14.8% in the brief counseling group and 10.5% in the assessment-only control group (Meyer et al., 2008). Further information regarding the ProGP study is described in Meyer et al. (2008). As the brief advice intervention was a mixture of oral advice by the general practitioner and written self-help materials, this study condition was excluded from the following data analyses. The baseline characteristics of the sample used in this study, including participants of the tailored letters intervention group and the assessment only group, are shown in Table 1. There were no significant differences between the two study groups with respect to these baseline characteristics. Data were collected at baseline and at 6-, 12-, 18-, and 24-month follow-ups. An additional assessment was performed after 3 months in the tailored letters condition for the purpose of the intervention. For the tailored letters group, the percentage of nonresponding participants at the follow-ups ranged from 31.8% (month 6) to 42.8% (month 24). For the assessment only group, the percentage of nonresponding participants at the follow-ups ranged from 21.2% (month 6) to 35.0% (month 24). No contact was the main reason for nonresponse. Participants who could not be recontacted at single follow-up assessments (n 524) differed from those who could be recontacted at each follow-up assessment (n 573) according to the variables sex, educational attainment, NFC, age, and number of cigarettes smoked per day ( p.05). Study Conditions Tailored-letters. In this treatment condition, patients received up to three individualized letters that were tailored according to the principles of the Transtheoretical Model of behavior change (TTM) (Velicer et al., 2000). Letters were tailored (a) to the stage of change, defined by current smoking status and intention to quit smoking, and (b) to scores on decisional balance, self-efficacy, and processes of change scales. The first letter, which was based on data gathered at the baseline assessment, included normative feedback, that is, feedback that depends on the individual scores compared to the population norm by stage, and was sent out within 1 week after the practice visit. The 3-month and 6-month letters included additional ipsative feedback, that is, feedback tailored to individual change according to the single constructs since the previous assessment. The tailored letters were exclusively text-based and their length varied between 2 and 3

3 NEED FOR COGNITION IN SMOKING CESSATION 369 Table 1 Baseline Characteristics of the Study Sample. Values are Numbers (Percentage) unless stated otherwise. Assessment only Tailored letters Total All subjects Gender Male 309 (50.7) 242 (49.6) 551 (50.2) Female 300 (49.3) 246 (50.4) 546 (49.8) Age, M (SD) 34.8 (13.4) 33.8 (13.2) 34.4 (13.3) Educational attainment No grade 22 (3.6) 15 (3.1) 37 (3.4) Secondary school 481 (79.0) 392 (80.3) 873 (79.6) Technical school 21 (3.4) 25 (5.1) 46 (4.2) High school 63 (10.3) 43 (8.8) 106 (9.7) No information 22 (3.6) 13 (2.7) 35 (3.2) Number of cigarettes smoked per day, M (SD) 16.2 (7.6) 16.4 (7.7) 16.3 (7.7) No information 4 (0.7) 4 (0.8) 8 (0.7) Stage of change Precontemplation 412 (67.7) 312 (63.9) 724 (66.0) Contemplation 173 (28.4) 148 (30.3) 321 (29.3) Preparation 14 (2.3) 14 (2.9) 28 (2.6) No information 10 (1.6) 14 (2.9) 24 (2.2) Smoking cessation self-efficacy, (scale 1 5) M (SD) 2.6 (0.9) 2.6 (0.8) 2.6 (0.9) No information 13 (2.1) 9 (1.8) 22 (2.0) 24-hour quit attempt during past year No attempt 426 (70.0) 340 (69.7) 766 (69.8) 1 attempt 183 (30.0) 146 (29.9) 329 (30.0) No information 0 (0.0) 2 (0.4) 2 (0.2) Need for cognition (scale 1 7), M (SD) 4.3 (0.9) 4.3 (0.8) 4.3 (0.9) No information 85 (14.0) 56 (11.5) 141 (12.9) pages. The letters were accompanied by a selection from a series of self-help manuals covering specific information relevant to the particular stage of change according to the TTM. Assessment only. No intervention, besides the usual routine treatment during a visit with a practitioner, was conducted on this group. No information about the participants was given to the practice team or the practitioner, and no self-help manuals were provided to the participants. Measurements For the baseline assessment, a questionnaire was administered in the waiting room of the general medical practice. Follow-up assessments were conducted 6, 12, 18, and 24 months after the initial practice visit via computer-assisted telephone interview. If an individual could not be reached by phone, a questionnaire was sent out. In the tailored-letters condition, an additional assessment was performed after 3 months for the purpose of the intervention. The follow-up assessments included items covering TTM constructs, smoking behavior, utilization of medical care and smoking cessation aids, and quality of life. NFC was measured at baseline using the German version (Bless, Wänke, Bohner, Fellhauer, & Schwarz, 1994) of the 18-item short form of the NFC Scale (Cacioppo, Petty, & Kao, 1984). Sample scale items included I would prefer simple to complex problems and It s enough for me that something gets the job done; I don t care how or why it works. Individuals indicated their agreement to each item on a 7-point rating scale (strongly agree strongly disagree). Higher values on the NFC scale represent a higher NFC. The original and the German version of this scale demonstrated good psychometric properties (Bless et al., 1994; Cacioppo et al., 1984). Cronbach s alpha for the current data set was.80, which is similar to that reported in a validation study for this short form of the German NFC scale (.83) (Bless et al., 1994), but lower than that of the original version (.90) (Cacioppo et al., 1984). Outcome variables. Smoking abstinence was assessed at 12-, 18-, and 24-month follow-ups using the criterion of 6 months prolonged abstention from smoking (Hughes et al., 2003). We used two questions to assess 6-month prolonged abstinence: First, participants were asked whether they were currently smoking cigarettes. They could choose one from three response options (a) Yes, I m smoking cigarettes daily, (b) Yes, I m smoking cigarettes occasionally, (c) No, I m not smoking cigarettes anymore. Second, participants who chose the third response option were asked whether they stopped cigarette smoking (a) within the last 6 months or (b) more than 6 months ago. Using Generalized Estimation Equation (GEE) analyses, the three point prevalence abstinence measures assessed at 12, 18, and 24 months were analyzed simultaneously. Smoking cessation self-efficacy was assessed with a 9-item scale (.95) (Jäkle, Keller, Baum, & Basler, 1999). Individuals indicated their confidence in refraining from smoking in different situations (e.g., I am confident I can resist smoking when I feel I need a lift ) on a 5-point Likert-type scale (1 - not at all confident and5-very confident). From the responses to the single items, a mean score for smoking cessation self-efficacy was calculated. Covariates. Stage of change was measured at baseline using the staging algorithm of the Transtheoretical Model (DiClemente et al., 1991). Individuals not willing to quit were allocated to the precon-

4 370 HAUG ET AL. templation stage; those considering quitting in the next 6 months and those intending to quit in the next 4 weeks (but not reporting a quit attempt in the past year), were classified in the contemplation stage. Smokers planning to quit in the next 4 weeks and having had a serious quit attempt in the last 12 months were allocated to the preparation stage. However, only a few persons were in the preparation stage of change at baseline. Therefore, we collapsed the participants from the contemplation and preparation stages into one category for the predictor and moderator analyses. Number of cigarettes smoked per day was assessed at baseline by the question: How many cigarettes do you smoke on an average day? Previous quit attempts were assessed by the question How many times in the last 12 months have you quit smoking cigarettes for at least 24 hours? This variable was dichotomized (0 vs. 1 or more quit attempts) for the current analyses. Educational attainment was assessed by a list of all educational levels that can be obtained in Germany. Participants could choose their highest educational level from this list. To allow an international comparability of the educational grades, we recategorized educational attainment into four categories: (a) no grade, (b) secondary school, (c) technical school, and (d) high school. As only a few persons were in the first and third category, we collapsed the first and second category (no grade or secondary school) as well as the third and the fourth category (technical or high school) for the predictor and moderator analyses. Data Analysis Because no associations between NFC and smoking related variables had been reported, we initially explored correlations of NFC with various smoking variables, which had been assessed at baseline. Concerning sociodemographic variables, previous studies had found a significant positive correlation between NFC and educational level, but no correlation of NFC with age or gender (Cacioppo et al., 1996). To replicate these results for our sample, and to identify potential covariates that should be considered in our multivariate prediction analyses, we examined correlations between NFC and sociodemographic variables. GEE analyses were carried out to investigate the relationship between baseline NFC, study group, and the outcome variables evaluated at various follow-up assessments over the two years of the study. GEE is a repeated-measures regression model that takes into account the correlation between the repeated measures within a person (Zeger, Liang, & Albert, 1988). We performed logistic GEE analyses for the binary outcome variable: smoking cessation; and linear GEE analysis for the continuous outcome variable: smoking cessation self-efficacy. To test NFC as a predictor, we used a multivariate prediction model for each outcome variable, controlling for (a) demographic variables that are associated with baseline NFC and (b) smoking variables that proved to be predictors of outcome in smoking cessation interventions. To test the independent predictive power of NFC, we simultaneously entered NFC, study group, educational level, number of cigarettes smoked per day, previous quit attempt, stage of change, and smoking cessation self-efficacy in our multivariate prediction models. Statistical analysis of moderators was performed following the guidelines of Kraemer, Wilson, Fairburn, and Agras (2006) for evaluating moderators of treatment-effects in randomized trials. Dichotomous baseline variables (including treatment condition) were coded as 0.5 and 0.5 and continuous baseline variables were centered at their mean. The GEE-models included the study group NFC interaction while controlling for the main effects of both study group and NFC. All analyses were performed using Stata, version 9.2. Given the clustered nature of the data (patients within general practices), we computed robust variance estimators for all GEE analyses. As the within-subject correlations for the outcome variables were of the same magnitude, we chose an exchangeable within-subject correlation structure for the GEE-models. An alpha level of 0.05 (2-tailed) was chosen for all statistical tests in this study. Multiple regression imputation methods were employed to impute single missing datapoints on baseline predictor variables, using the Stata ICE procedure (Royston, 2005). We applied the intention to treat principle and considered all randomized subjects in the analyses. Missing data on the outcome variable 6-month prolonged smoking abstinence were coded as smokers. For participants with missing data on smoking cessation selfefficacy at any of the follow-ups, we carried forward the baseline data. Using these imputed data, significant moderator effects were illustrated by plotting the outcome variable for persons with high versus low baseline NFC (median split) separately for the two study groups. Results Correlations of NFC with Demographic and Smoking Variables Baseline NFC was positively correlated with higher educational attainment (Kendall s.25, p.01), but it was not correlated with age (r s.01, p.96) nor gender (r pb.05, p.09). NFC was only marginally, but significantly correlated with smoking cessation self-efficacy (r s.12, p.01). There were no significant correlations of NFC with number of cigarettes smoked per day (r s.00, p.92), with the occurrence of a quit attempt in the previous year (r pb.01, p.81), nor with baseline stage of change (Kendall s.05, p.03). NFC as Predictor of Outcome The multivariate prediction models, controlling for study group, educational attainment, and the baseline smoking variables stage of change, smoking cessation self-efficacy, cigarettes per day and a previous quit attempt, revealed that NFC successfully predicted 6-month smoking abstinence (odds ratio 1.45; p.01) (see Table 2) and also smoking cessation self-efficacy ( p.01) (see Table 3). NFC as Moderator of Outcome When compared to the assessment-only condition, NFC did not moderate the effect of the tailored letters intervention on 6-month smoking abstinence (.28, SE.28, t 1.03, p.30) but on smoking cessation self-efficacy (.13, SE.06, t 2.01, p.05). Specifically, tailored letters resulted in higher smoking cessation self-efficacy only for participants with high NFC. For persons with low NFC, self-efficacy was comparable to persons in the assessment-only group (see Figure 1).

5 NEED FOR COGNITION IN SMOKING CESSATION 371 Table 2 Multiple Prediction Model of 6-Month Prolonged Smoking Abstinence Variable SE t p OR (95% CI) Study group assessment only (ref.) Study group tailored letters ( ) Time of measurement (month) ( ) No grade or secondary school (ref.) Technical or high school ( ) Precontemplation stage of change at baseline (ref.) Contemplation/preparation stage of change at baseline ( ) Smoking cessation self-efficacy at baseline ( ) Cigarettes smoked per day at baseline ( ) Quit attempt during past year at baseline ( ) Need for cognition at baseline ( ) Note. Logistic generalized estimation equation model. Six-month prolonged smoking abstinence was assessed at 12-, 18-, and 24-month follow-ups. OR odds ratio; CI confidence interval. Discussion This study tested NFC as a predictor and a moderator of outcome in a written tailored letters smoking cessation intervention. The results showed that NFC is an independent predictor of both smoking cessation and smoking cessation self-efficacy. Higher values of baseline NFC resulted in higher abstinence rates and higher smoking cessation self-efficacy at the follow-up assessments. These results confirmed our hypothesis that NFC constitutes an independent predictor of smoking cessation and smoking cessation self-efficacy. Irrespective of whether the patients received computer tailored-letters or assessment only, higher baseline NFC resulted in higher smoking cessation and smoking cessation self-efficacy at the follow-up assessments. Concerning smoking cessation self-efficacy, this result is consistent with results from a previous study (Steward et al., 2003), which showed that baseline NFC was predictive of self-efficacy to quit smoking. Our data did not confirm the hypothesis that NFC moderates the effectiveness of the tailored letters intervention compared to assessment only when using smoking abstinence as outcome. However, NFC moderated the effect of the tailored letters intervention on smoking cessation self-efficacy when compared to the assessment-only condition. The limitation that we did not biochemically verify smoking status at follow-up assessments should be considered. However, because of the low demand characteristics of our study (low intensity treatments, limited face-to-face contact) we do not expect the lack of biochemical validation to have resulted in any strong bias in our data (Velicer et al., 1992). Our results have two important implications. First, our findings suggest that the effectiveness of written interventions for smoking cessation could be enhanced by also considering NFC as tailoring variable. The intervention in this study primarily relied on facts that were provided in written format. As a result, this intervention was primarily matched to persons with high NFC. Approaches that might be more effective for persons with low NFC could be: to use more emotion based, evaluative messages (Vidrine et al., 2007), to integrate more pictographic information (Carnaghi, Cadinu, Castelli, Kiesner, & Bragantini, 2007), or to provide shorter but more repetitive interventions (e.g., via mobile phone text messaging; Haug et al., 2008; Rodgers et al., 2005). However, it has to be taken into account that increasing the number of items to assess additional tailoring variables might be time-consuming and it might impede the feasibility and acceptance of brief interventions. This applies particularly for the current NFC scale, consisting of Table 3 Multiple Prediction Model of Smoking Cessation Self-Efficacy Variable SE t p Study group assessment only (ref.) Study group tailored letters Time of measurement (month) No grade or secondary school (ref.) Technical or high school Precontemplation stage of change at baseline (ref.) Contemplation/preparation stage of change at baseline Smoking cessation self-efficacy at baseline Cigarettes smoked per day at baseline Quit attempt during past year at baseline Need for cognition at baseline Note. Linear generalized estimation equation model. Smoking cessation self efficacy was assessed at 6- and 24-month follow-ups.

6 372 HAUG ET AL. Figure 1. Need for cognition (NFC) moderating the effect of the tailored letters intervention (TL) compared to assessment only (AO) on smoking cessation self-efficacy. 18 items. Because of its association with educational attainment, which could be assessed using a single item, its practicability as well as its effectiveness to enhance tailored interventions is still ambiguous and should be tested in future studies. Second, the results from the moderator analyses indicate that NFC had the strongest impact on smoking cessation self-efficacy within the tailored letters intervention. The graphical illustration of this moderator effect (see Figure 1) shows that tailored letters resulted in a substantial increase in smoking cessation self-efficacy only for participants with high NFC. In contrast, for persons with low NFC, the results with tailored letters were similar to those of the assessment-only condition. This suggests that higher levels of NFC are required to increase smoking cessation self-efficacy in the tailored letters intervention. This study provides preliminary evidence that matching smoking cessation interventions to an individual s NFC might be useful to improve their effectiveness. However, to obtain more conclusive results, the effectiveness of interventions that are tailored to an individual s NFC compared to interventions that are not, should be tested in prospective studies. References Abrams, D. B., Herzog, T. A., Emmons, K. M., & Linnan, L. (2000). Stages of change versus addiction: A replication and extension. Nicotine and Tobacco Research, 2, Bless, H., Wänke, M., Bohner, G., Fellhauer, R. F., & Schwarz, N. (1994). Need for cognition: Eine Skala zur Erfassung von Engagement und Freude bei Denkaufgaben [Need for cognition: A scale measuring engagement and happiness in cognitive tasks]. Zeitschrift für Sozialpsychologie, 25, Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, Cacioppo, J. T., Petty, R. E., Feinstein, J. A., & Jarvis, W. B. G. (1996). Dispositional differences in cognitive motivation: The life and time of individuals varying in need for cognition. Psychological Bulletin, 119, Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48, Carnaghi, A., Cadinu, M., Castelli, L., Kiesner, J., & Bragantini, C. (2007). The best way to tell you to use a condom: The interplay between message format and individuals level of need for cognition. AIDS Care, 19, DiClemente, C. C., Prochaska, J. O., Fairhurst, S. K., Velicer, W. F., Velasquez, M. M., & 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, Farkas, A. J., Pierce, J. P., Zhu, S. H., Rosbrook, B., Gilpin, E. A., Berry, C., & Kaplan, R. M. (1996). Addiction versus stages of change models in predicting smoking cessation. Addiction, 91, ; discussion Haug, S., Meyer, C., Gross, B., Schorr, G., Thyrian, J. R., Kordy, H.,... John, U. (2008). Kontinuierliche individuelle Förderung der Rauchabstinenz bei sozial benachteiligten jungen Erwachsenen über das Handy Ergebnisse einer Pilotstudie [Continuous individual support of smoking cessation in socially deprived young adults via mobile phones - results of a pilot study]. Gesundheitswesen, 70, Hughes, J. R., Keely, J. P., Niaura, R. S., Ossip-Klein, D. J., Richmond, R. L., & Swan, G. E. (2003). Measures of abstinence in clinical trials: Issues and recommendations. Nicotine and Tobacco Research, 5, Hymowitz, N., Cummings, K. M., Hyland, A., Lynn, W. R., Pechacek, T. F., & Hartwell, T. D. (1997). Predictors of smoking cessation in a cohort of adult smokers followed for five years. Tobacco Control, 6(Suppl. 2), Jäkle, C., Keller, S., Baum, E., & Basler, H.-D. (1999). Skalen zur Selbstwirksamkeit und Entscheidungsbalance im Prozeß der Verhaltensänderung von Rauchern [Scales for the measurement of self-efficacy and decisional balance in the process of behavioral change in smokers]. Diagnostica, 45, Kraemer, H. C., Wilson, G. T., Fairburn, C. G., & Agras, W. S. (2006). Mediators and moderators of treatment effects in randomized clinical trials. Archives of General Psychiatry, 59, Kreuter, M. W., Bull, F. C., Clark, E. M., & Oswald, D. L. (1999). Understanding how people process health information: A comparison of tailored and nontailored weight-loss materials. Health Psychology, 18, Lancaster, T., & Stead, L. F. (2005). Self-help interventions for smoking cessation. Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No.: CD DOI: / CD pub2 Meyer, C., Ulbricht, S., Baumeister, S. E., Schumann, A., Rüge, J., Bischof, G.,... John, U. (2008). Proactive interventions for smoking cessation in general medical practice: A quasi-randomized controlled trial to examine the efficacy of computer-tailored letters and physiciandelivered brief advice. Addiction, 103, Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review of tailored print health behavior change interventions. Psychological Bulletin, 133, Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L. Berkowitz (Ed.), Advances in experimental and social psychology (Vol. 19, pp ). New York: Academic Press. Petty, R. E., Haugtvedt, C. P., & Smith, S. M. (1995). Elaboration as a determinant of attitude strength: Creating attitudes that are persistent, resistant, and predictive of behavior. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp ). Mahwah, NJ: Erlbaum. Prochaska, J. O., & Velicer, W. F. (1997). The transtheoretical model. American Journal of Health Promotion, 12, 6 7. Rodgers, A., Corbett, T., Bramley, D., Riddell, T., Wills, M., Lin, R. B.,... Jones, M. (2005). Do u smoke after txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tobacco Control, 14, Royston, P. (2005). Multiple imputation of missing values: Update of ICE. The Stata Journal, 5, Steward, W. T., Schneider, T. R., Pizarro, J., & Salovey, P. (2003). Need for cognition moderates responses to framed smoking-cessation messages. Journal of Applied Social Psychology, 33,

7 NEED FOR COGNITION IN SMOKING CESSATION 373 Velicer, W. F., Prochaska, J. O., Fava, J. L., Rossi, J. S., Redding, C. A., Laforge, R. G.,... Robbins, M. L. (2000). Using the transtheoretical model for population-based approaches to health promotion and disease prevention. Homeostasis, 40, Velicer, W. F., Redding, C. A., Sun, X., & Prochaska, J. O. (2007). Demographic variables, smoking variables, and outcome across five studies. Health Psychology, 26, Vidrine, J. I., Simmons, V. N., & Brandon, T. H. (2007). Construction of smoking-relevant risk perceptions among college students: The influence of need for cognition and message content. Journal of Applied Social Psychology, 37, Zeger, S. L., Liang, K. Y., & Albert, P. S. (1988). Models for longitudinal data: A generalized estimating equation approach. Biometrics, 44,