ABSTRACT

Objective: To develop a new scale aimed at evaluating smoking motivation by incorporating questions and domains from the 68-item Wisconsin Inventory of Smoking Dependence Motives (WISDM-68) into the Modified Reasons for Smoking Scale (MRSS). Methods: Nine WISDM-68 questions regarding affiliative attachment, cue exposure/associative processes, and weight control were added to the 21 questions of the MRSS. The new scale, together with the Fagerström Test for Nicotine Dependence (FTND), was administered to 311 smokers (214 males; mean age = 37.6 ± 10.8 years; mean number of cigarettes smoked per day = 15.0 ± 9.2), who also provided additional information. We used exploratory factor analysis in order to determine the factor structure of the scale. The influence that certain clinical features had on the scores of the final factor solution was also analyzed. Results: The factor analysis revealed a 21-question solution grouped into nine factors: addiction, pleasure from smoking, tension reduction, stimulation, automatism, handling, social smoking, weight control, and affiliative attachment. For the overall scale, the Cronbach's alpha coefficient was 0.83. Females scored significantly higher for addiction, tension reduction, handling, weight control, and affiliative attachment than did males. The FTND score correlated positively with addiction, tension reduction, stimulation, automatism, social smoking, and affiliative attachment. The number of cigarettes smoked per day was associated with addiction, tension reduction, stimulation, automatism, affiliative attachment, and handling. The level of exhaled CO correlated positively with addiction, automatism, and affiliative attachment. Conclusions: The new scale provides an acceptable framework of motivational factors for smoking, with satisfactory psychometric properties and reliability.

Keywords: Smoking; Validation studies; Tobacco use cessation.

RESUMO

Objetivo: Desenvolver uma nova escala voltada para a avaliação da motivação para fumar, incorporando questões do 68-item Wisconsin Inventory of Smoking Dependence Motives (WISDM-68, Inventário Wisconsin dos Motivos de Dependência ao Fumo, de 68 itens) na Modified Reasons for Smoking Scale (MRSS. Escala Razões para Fumar Modificada). Métodos: Nove questões do WISDM-68 relativas à associação estreita, exposição a gatilhos/processos associativos e controle de peso foram incorporadas às 21 questões da MRSS. Um total de 311 fumantes (214homens; idade média = 37,6 ± 10,8 anos; média de cigarros consumidos ao dia = 15,0 ± 9,2) responderam a nova escala, o Fagerström Test for Nicotine Dependence (FTND, Teste de Fagerström para Dependência de Nicotina) e outras questões. Empregamos a análise fatorial exploratória para determinar a estrutura fatorial da escala. A influência de algumas características clínicas nos escores da solução fatorial final foi também avaliada. Resultados: A análise fatorial revelou uma solução com 21 questões agrupadas em nove fatores: dependência, prazer de fumar, redução da tensão, estimulação, automatismo, manuseio, tabagismo social, controle de peso e associação estreita. Para a escala como um todo, o coeficiente alfa de Cronbach foi de 0,83. As mulheres exibiram maiores escores para dependência, redução da tensão, manuseio, controle de peso e associação estreita do que os homens. Os escores do FTND correlacionaram-se positivamente com dependência, redução da tensão, estimulação, automatismo, tabagismo social e associação estreita. O número de cigarros fumados ao dia se associou com dependência, redução da tensão, estimulação, automatismo, associação estreita e manuseio. Os níveis de CO exalado mostraram associações positivas com dependência, automatismo e associação estreita. Conclusões: A nova escala fornece um quadro aceitável dos fatores motivacionais associados ao tabagismo, com confiabilidade e propriedades psicométricas satisfatórias.

Palavras-chave: Tabagismo; Estudos de validação como assunto; Abandono do uso de tabaco.

Introduction

The psychoactive properties of nicotine are generally considered to constitute the most important element of nicotine dependence. However, there is evidence that the motivation for smoking is multidimensional.(1-3)

Therefore, nicotine addiction might involve factors other than physical dependence on nicotine, which would mean that nicotine dependence is a broad construct, comprising various psychosocial facets. Accurate identification of the distinctive factors that drive people to smoke might inform decisions regarding the development of public policies for smoking prevention and control, as well as those regarding the design of tailored smoking cessation strategies.

The motivations for smoking have been studied for decades, and some motivational factors have been described, including the desire to maximize positive affects/minimize negative affects, addiction, and habit.(2) Based on this model, Horn & Waingrow created the Reasons for Smoking Scale (RSS), which has long been the instrument most commonly employed to measure smoking motivation in North America.(4-6)

In 1969, the RSS was applied to 2,094 adult smokers, and six motivational elements were identified by factor analysis: stimulation; pleasurable relaxation; habitual smoking; addictive smoking; negative affect reduction; and sensorimotor manipulation.(7) Other authors have studied the RSS and have obtained comparable results.(8)

The authors of a study conducted in France suggested a change in the RSS, introducing three new items related to an additional motivational factor designated "social smoking" and bringing the total number of items to 21. The new instrument was designated the Modified Reasons for Smoking Scale (MRSS).(9) The psychometric properties of this scale were evaluated in a group of 330 adult smokers, and seven factors were identified: addictive smoking; pleasure from smoking; tension reduction/relaxation; social smoking; stimulation; habit/automatism; and handling. Two items exhibited low factor loadings and were excluded from the final version of the scale, which was therefore composed of 19 questions.(9)

Another tool that was developed to characterize the motivational reasons for smoking is the 68-item Wisconsin Inventory of Smoking Dependence Motives (WISDM-68).(10,11) This instrument is based on theoretically grounded motives for drug use, approaches previously uninvestigated aspects, and can be considered the most complete evaluation of smoking motivation available. A study involving 775 smokers showed that the subscales of the WISDM-68 have acceptable internal consistency, are differentially present across the levels of smoking intensity, and have a multidimensional structure.(10) In addition, validity analyses indicated that the subscales are significantly related to the intensity of smoking, as well as to symptoms of dependence and relapse, as defined in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.(11) The final version of the WISDM-68 includes 68 items that assess 13 motives for smoking: affiliative attachment; automaticity; loss of control; behavioral choice/melioration; cognitive enhancement; craving; cue exposure/associative processes; negative reinforcement; positive reinforcement; social/environmental goads; taste/sensory process; tolerance; and weight control. The number of items related to each motive ranges from 4 to 7, most motives comprising 5 items.

The WISDM-68 provides a comprehensive evaluation of smoking motivations. However, the great number of questions makes its routine use in clinical practice somewhat problematic. In addition, a close comparison between the questions and the factors derived from the MRSS and the WISDM-68 reveals some degree of overlap between the solutions obtained. Of the 13 WISDM-68 motives, 9 are represented, to some degree, in the MRSS.

The WISDM-68 automaticity motive, for example, is defined by 5 items: "I often smoke without thinking about it"; "I smoke without deciding to"; "I frequently light cigarettes without thinking about it"; "I find myself reaching for cigarettes without thinking about it"; and "Sometimes I'm not aware that I'm smoking". Similarly, the MRSS habit/automatism motive is defined by the following phrases: "I smoke cigarettes automatically without even being aware of it"; "I light up a cigarette without realizing I still have one burning in the ashtray"; and "I've found a cigarette in my mouth and did not remember putting it there".

Other examples of similarities between the elements of the two scales are the WISDM-68 cognitive enhancement motive and the MRSS stimulation motive. The former is defined by the following sentences: "I smoke when I really need to concentrate"; "I frequently smoke to keep my mind focused"; "Smoking helps me stay focused"; "My concentration is improved after smoking a cigarette"; and "Smoking helps me think better". The latter is defined by the following sentences: "I smoke cigarettes to keep myself from slowing down"; "I smoke cigarettes to stimulate me, to perk myself up"; and "I smoke cigarettes to give me a lift". However, 4 of the 13 WISDM-68 motives appear not to be minimally covered by the MRSS: affiliative attachment; behavioral choice/melioration; cue exposure/associative processes; and weight control.

The objective of the present study was to describe the development, the psychometric properties, and the factor structure of a new questionnaire, containing the seven traditional factors of the MRSS and certain items derived from the WISDM-68.

The final instrument, designated the University of São Paulo Reasons for Smoking Scale (USP-RSS), has the potential to become a relevant tool for the evaluation of smoking motivations in various clinical settings.

Methods

The new instrument was developed simultaneously with the analysis of the factor structure and psychometric properties of a Brazilian Portuguese-language version of the MRSS.(12) Volunteers (smokers only) were recruited from among consecutive blood donors at the Ribeirão Preto Blood Bank, located in the city of Ribeirão Preto, Brazil. All volunteers completed a single form containing 53 questions. In all cases, the volunteers were able to complete the form in less than 60 min. The data collected were employed in order to analyze the structure and properties of the Brazilian Portuguese-language version of the MRSS, as well as to investigate the effect of the addition of 9 questions derived from the WISDM-68. Although the data had been obtained on a single occasion, due to the distinctive objectives of the investigations, the individual statistical analysis, and the enormous amount of results, we chose to report our results in two different studies, one by Souza et al.(12) and the present one.

The original MRSS scale, in English, was kindly provided by Ivan Berlin, of the Centre Hospitalier Universitaire Pitié-Salpêtrière (Paris, France). The WISDM-68 questionnaire had been previously published, and electronic consent for its use was obtained from the first author of the original article, Megan E. Pipe, of the Center for Tobacco Research and Intervention, at the University of Wisconsin Medical School (Madison, WI, USA).

The WISDM-68 questions dealing with motives that are not covered by the MRSS (affiliative attachment; cue exposure/associative processes; weight control; and behavioral choice/melioration) were considered for addition to the MRSS. In order to minimize the number of items, only 3 representative questions from each of the first 3 motives were chosen to be inserted into the new instrument. The authors deemed that the behavioral choice/melioration motive would add little to the potential clinical usefulness of the new scale, and questions related to this domain were not included. The scoring system used in the MRSS, a Likert scale ranging from 1 to 5, was applied to all 30 questions. The original composition of the new questionnaire is available in the online version of this publication (Appendix 1).

All 30 questions were submitted to a process of translation to Brazilian Portuguese and transcultural adaptation for use in Brazil, as previously described.(13) A final, consensus Brazilian Portuguese-language version was employed in the subsequent studies regarding factor structure, reliability, and concurrent validity.

Blood donors reporting to have smoked at least one cigarette a day in the last week were invited to participate in the study. Subjects under 18 years of age, reporting clinical or psychiatric comorbidities, or having a history of illicit drug addiction/alcoholism were excluded, as were those who were illiterate individuals or whose native language was not Portuguese. The data derived from this group of smokers were employed in the investigation of the factor structure and in the concurrent validation of the instrument.

Seated in a quiet environment, all of the volunteers completed a standardized questionnaire including the items of the new scale, together with the Fagerström Test for Nicotine Dependence (FTND), as well as providing information regarding their smoking history, marital status, and level of education.(14) The subjects were subsequently asked to exhale into a CO monitor (Micro CO, Micro Medical Ltd, Rochester, England).
A second group of smokers, recruited from among the employees of the University of São Paulo at Ribeirão Preto School of Medicine Hospital das Clínicas, in Ribeirão Preto, Brazil, were selected for the analysis of test-retest reliability and completed the same questionnaire twice, 15 days apart.

The project design was approved by the local ethics on research committee, and it followed, during its development, the principles set forth in the Declaration of Helsinki. All participating volunteers gave written informed consent.

The results were analyzed with the Statistical Package for the Social Sciences, version 13.0 (SPSS Inc., Chicago, IL, USA). Demographic data, smoking history data, and the scores obtained are expressed as means and standard deviations.

The scores on the 30 questions concerning smoking motivation were evaluated by exploratory factor analysis, using main component analysis and oblique rotation, aiming at the construct validation of the new scale.(15) The strategy used for factor extraction was principal axis factoring, because this method extracts most of the variance of the variables and has been traditionally used in order to reduce great amounts of data to smaller, concise sets of variables. The selection of the extracted factors was performed by applying the Kaiser criterion, which supposes that factors with eigenvalues < 1 should not be retained.(16) In addition, only the items with factor loadings > 0.4 were selected as components of the extracted factors. The internal consistency of the factors generated was assessed with Cronbach's alpha coefficient.(17) The test-retest reliability of the final version of the scale was evaluated using weighted kappa statistics and by determination of intraclass correlation coefficients (ICCs).(18,19)

The influence of individual clinical features on the factor set as a whole was analyzed with multivariate ANOVA and Pearson's correlation coefficient.(20) The independent variables gender and marital status were classified as categories, and the variables age, years of schooling, smoking duration, cigarettes per day, FTND score, and exhaled CO level were considered continuous cofactors. When statistically significant differences were detected, univariate ANOVA was performed between each factor and the specific independent variable. The level of significance was set at p ≤ 0.05.

Results

The study sample consisted of 311 smokers with a mean age of 37.6 ± 10.8 years. There was a predominance of males (214/68.8%), and the degree of nicotine addiction could be classified, on average, as low (FTND score = 3.7 ± 2.4; number of cigarettes smoked per day = 15.0 ± 9.2). Most of the subjects (56.6%) smoked < 20 cigarettes per day, 39.5% smoked 20-40 cigarettes per day, and 3.9% smoked > 40 cigarettes per day. The mean number of years of schooling was 9.0 ± 3.8, and 212 subjects (68.2%) reported having a steady partner.

The ratio between the number of subjects and the number of items in the scale was 10.37, which was considered acceptable for the factor analysis.

The exploratory factor analysis assumptions were tested. The Kaiser-Meyer-Olkin index of sampling adequacy was 0.78, which can be described as being between meritorious and satisfactory. Bartlett's sphericity test rejected the hypothesis that the correlation matrix was an identity matrix (p < 0.001; χ2 = 2,603.8).

The total factor variance and the eigenvalues obtained by factor analysis are reported in Table 1. Ten factors explained 62.5% of the total variation.



The composition of those ten factors and the loading values for each item are reported in Table 2. The application of 0.4 as the cut-off loading value for the inclusion of items into the factors led to the exclusion of 9 questions (numbers 3, 8, 10, 14, 16, 22, 23, 29, and 30). The initial composition of the factors and their respective items were as follows: Factor 1-items 6 and 26; Factor 2-items 1, 11, and 20; Factor 3-items 4 and 13; Factor 4-items 7 and 21; Factor 5-items 5, 15, and 25; Factor 6-items 17 and 28; Factor 7-items 9 and 19; Factor 8-items 2 and 12; Factor 9-items 18 and 27; and Factor 10-item 24.



Because factor 10 (cue exposure/associative processes) was composed of one item only, we decided to remove it from the new scale. The evaluation of the internal consistency for the remaining nine factors revealed the following Cronbach's alpha coefficient values: Factor 1-α = 0.65; Factor 2-α = 0.77; Factor 3-α = 0.82; Factor 4-α = 0.88; Factor 5-α = 0.75; Factor 6-α = 0.64; Factor 7-α = 0.72; Factor 8-α = 0.76; and Factor 9-α = 0.47.

Because the Cronbach's alpha coefficient for factor 9 was low, a decision was made to reintroduce question 8, which had shown a factor weight of 0.333, into this factor, for the purpose of achieving better internal consistency. The Cronbach's alpha coefficients for this new factor 9 and for the overall scale were, respectively, 0.54 and 0.83. The final composition of the new scale, with the appropriate denominations for the factors and the selected items (questions), was as follows: Factor 1-addiction (questions 6 and 26); Factor 2-stimulation (questions 1, 11, and 20); Factor 3-pleasure from smoking (questions 4 and 13); Factor 4-affiliative attachment (questions 7 and 21); Factor 5-tension reduction (questions 5, 15, and 25); Factor 6-weight control (questions 17 and 28); Factor 7-social smoking (questions 9 and 19); Factor 8-handling (questions 2 and 12); and Factor 9-automatism (questions 8, 18, and 27). The mean values and the respective standard deviations of the scores observed in the sample of 311 volunteers are illustrated in Figure 1.



The test-retest reliability of the new scale was evaluated in a second group, composed of 54 smoking volunteers (mean age: 41.3 ± 10.9 years; FTND score: 4.3 ± 2.7; 19 males). The participants in this group completed the questionnaire twice, 15 days apart. The weighted kappa coefficients for the items in the final version of the scale were significant (p < 0.01 for all; Table 3).



The ICCs for the nine generated factors were as follows: addiction-ICC = 0.769; stimulation-ICC = 0.763; pleasure from smoking-ICC = 0.618; affiliative attachment-ICC = 0.796; tension reduction-ICC = 0.802; weight control-ICC = 0.864; social smoking-ICC = 0.823; handling-ICC = 0.798; and automatism-ICC = 0.851.

The influence of the clinical characteristics of the 311 smokers on the scores of the nine detected factors is demonstrated in Table 4. Females exhibited significantly higher scores for addiction, tension reduction, handling, weight control, and affiliative attachment. In addition, females showed a trend toward higher scores for social smoking.



All of the features related to smoking history were significantly associated with at least some of the motivational factors. The FTND scores correlated positively with addiction, tension reduction, stimulation, automatism, social smoking, and affiliative attachment. Among the motivational factors, addiction correlated most strongly with the FTND score (r = 0.50, p < 0.001). The number of cigarettes smoked per day correlated with addiction, tension reduction, stimulation, automatism, affiliative attachment, and handling. Smoking history was positively associated with automatism and affiliative attachment. The level of exhaled CO, measured just after the questionnaire had been completed, correlated positively with addiction, automatism, and affiliative attachment.

The level of education was positively associated with pleasure from smoking and was negatively associated, albeit weakly so, with stimulation.
We observed a trend toward negative associations between the following: age and addiction; age and pleasure from smoking; and age and tension reduction. Marital status had no influence on any of the factor scores.

Discussion

This study reports the adaptation of a previously validated clinical instrument for the evaluation of smoking motivation, the MRSS. Our purpose was to obtain a new measurement tool that would provide adequate coverage of constructs while maintaining brevity in terms of its administration. The proposed modification was the introduction of three new domains derived from a comprehensive and more modern investigation. Among the intended changes, only two factors revealed well characterized properties that would justify their inclusion into the new tool.

The instrument derived, designated the USP-RSS, exhibited satisfactory factor structure and preliminary psychometric properties. The final Brazilian Portuguese-language version of the USP-RSS is available (as Appendix 2) in the online version of this article.

Since 1966, the RSS has been commonly used to measure the motivations for smoking. There is a substantial amount of data in the literature supporting this particular smoking typology.(4,8,21) One review of the literature concluded that the RSS has stable factor structure, internal consistency, and temporal stability.(8) However, according to the authors of that review, there is as yet insufficient evidence of its validity. The RSS motivational profile of smokers was developed more than four decades ago, and the reasons that led people to smoke might have changed during this time. The WISDM-68 was recently developed as the result of a distinct methodology and provides an updated solution with 13 motives for smoking. Nevertheless, a close observation of the results suggests that the WISDM-68 identifies a substantial number of factors already present, to some degree, in the traditional RSS and in its modified version, the MRSS. In view of this, the truly innovative motives of the WISDM-68 would be the following: affiliative attachment; cue exposure/associative processes; weight control; and behavioral choice/melioration.

In order to generate a more comprehensive tool, suitable for use in routine clinical practice, 9 questions related to 3 motives (affiliative attachment; cue exposure/associative processes; and weight control) were incorporated into the original 21-item version of the MRSS. An effort was made to develop a simple instrument, and the original scoring system of the MRSS was applied to all of the questions.

The factor analysis initially revealed a 10-item solution. In order to optimize the properties of the final scale, only nine factors comprising at least 2 questions with meaningful factor loadings were selected. The internal consistency for most of the defined factors was acceptable, the Cronbach's alpha coefficients being > 0.7 for six of them. Although the Cronbach's alpha coefficient for automatism was low, even with the inclusion of question 8, we opted to retain it as part of the instrument due to its potential clinical importance, as previously described.(9,22) In addition, the overall Cronbach's alpha coefficient of 0.83 for the nine-factor scale can be considered satisfactory.

The test-retest reliability of the final version of the scale was evaluated in a distinct group of 54 smokers, all of whom completed the scale on two occasions, 15 days apart. The weighted kappa coefficients and the ICCs indicated that the USP-RSS exhibits good temporal consistency.

The motivation profile of the primary study group (311 smokers) featured high mean scores for addiction, pleasure from smoking, and tension reduction; intermediate scores for affiliative attachment, social smoking, and handling; and low scores for stimulation, weight control, and automatism (Figure 1). In a previous study, the scores for addiction, pleasure from smoking, and tension reduction were also high among smokers, although the degree of nicotine dependence was higher among the smokers evaluated in that study than among those in our sample.(9)
These results strongly suggest that these three factors are central elements in the development and persistence of nicotine addiction.

Several factors of the USP-RSS showed significant associations with measurements of smoking intensity. For instance, addiction correlated positively with the number of cigarettes smoked per day, the FTND score, and the level of exhaled CO. These results indicate that addiction is also a satisfactory proxy for the physical dependence of smokers. The FTND score and the number of cigarettes smoked per day also correlated significantly with tension reduction, stimulation, automatism, and affiliative attachment. The finding that the level of exhaled CO correlated significantly with automatism and affiliative attachment is also noteworthy. These results indicate that the final version of the proposed instrument exhibits adequate concurrent validity.

Automatism and affiliative attachment were significantly associated with a greater number of measurements of smoking intensity than were any of the other motives. This suggests that these two motives play important roles in the persistence of smoking dependence.

Automatism is related to smoking without intention or awareness. This factor also correlated significantly with the number of cigarettes smoked per day, and its importance in smoking dependence has been highlighted by other authors.(8,9,22) In another study, multivariate logistic regression revealed that higher automatism scores were predictive of failure to quit smoking.(9) Automatism may be secondary to conditioned behaviors and repetitive rewarding actions, as well as contributing to the lack of control of smoking.

Affiliative attachment can be defined as a strong emotional connection to smoking and cigarettes. Subjects with a history of long and heavy smoking exposure are most likely at the greatest risk of developing an emotional attachment to cigarettes. It is also possible that this dimension is more prevalent in smokers with psychosocial problems. High scores for this motivational domain might reflect nicotine dependence that is more complex. It remains to be seen if this factor is going to influence the results of smoking cessation interventions.

Apparently, handling and social smoking are only weakly connected with smoking intensity, whereas pleasure from smoking and weight control are not related to it at all. We found that gender exerted a significant influence on several motivational scores. In our study, females more often smoked due to physical dependence (addiction), tension reduction, handling, weight control, and affiliative attachment. These findings are relevant, because they might help to explain why female smokers find it more difficult to quit and feel more dependent on cigarettes than do men.(23)

The high scores for addiction found here confirm previous reports of lower smoking cessation rates among women under nicotine replacement therapy.(23-25) In addition, weight control has been described as a strong motivation for women to smoke.(26,27) It has been shown that smoking initiation is higher among adolescent girls who report elevated perception of the importance of being thin.(27) It is also of note that, in the present study, there was a trend toward higher social smoking scores among women, social smoking having previously been described as a relevant motivator of female smoking.(9)

Among the limitations of this study, the mean degree of nicotine dependence in this group of smokers, as evaluated by the FTND, can be classified as low. This finding probably reflects the real smoking profile observed in the community, rather than that seen in specialized smoking cessation clinics. Indeed, the proportion of smokers in Brazil has been reported to have decreased progressively over recent decades, and it would not be totally unexpected to observe a similar trend in the severity of the addiction.

Because the results obtained strongly reflect the composition of the original elements of the MRSS and WISDM-68, it is unlikely that the application of the scale to a group of heavier smokers would produce a different factor solution. However, it is possible that the answers of heavy smokers could lead to different factor loadings, and the subsequent inclusion of additional items in some of the subscales. Another major limitation of this study is that the volunteers were neither enrolled in a smoking cessation program nor belonged to populations at high risk of heavy smoking, such as psychotic or drug-addicted subjects. Further studies will be needed in order to determine the applicability of this scale in such groups. The present study involved Brazilian Portuguese-speaking smokers, employing questions that had been translated from the language of the original instruments. Therefore, linguistic and cultural differences might have influenced the final results. Authors interested in employing this scale in speakers of languages other than Portuguese are advised to develop their own factor and psychometric analysis, on the basis of the model presented in Appendix 1.

In conclusion, the USP-RSS provides a distinct framework of motivational factors for smoking, with good reliability and some satisfactory psychometric properties. Additional studies are still needed in order to evaluate the full validity and the true usefulness of the USP-RSS in smoking cessation interventions.

The final Portuguese-language version of the USP-RSS is available on line at www.jornaldepneumologia.com.br



References


1. Carmo JT, Andrés-Pueyo A, López EA. The evolution in the concept of smoking [Article in Spanish]. Cad Saude Publica. 2005;21(4):999-1005.

2. Tomkins SS. Psychological model for smoking behavior. Am J Public
Health Nations Health. 1966;56(12):Suppl 56:17-20.

3. Zbikowski SM, Swan GE, McClure JB. Cigarette smoking and nicotine dependence. Med Clin North Am. 2004;88(6):1453-65, x.

4. Currie SR. Confirmatory factor analysis of the Reasons for Smoking Scale in alcoholics. Nicotine Tob Res. 2004;6(3):465-70.

5. Horn D, Waingrow S. Behavior and attitudes questionnaire. Bethesda: National Clearinghouse for Smoking and Health; 1966.

6. Sánchez-Johnsen L, Ahluwalia JS, Fitzgibbon M, Spring BJ. Ethnic similarities and differences in reasons for smoking. Addict Behav. 2006;31(3):544-8.

7. Ikard FF, Green DE, Horn D. A scale to differentiate between types of smoking as related to the management of affect. Int J Addict. 1969;4(4):649-59.

8. Tate JC, Schmitz JM, Stanton AL. A critical review of the Reasons for Smoking Scale. J Subst Abuse. 1991;3(4):441-55.

9. Berlin I, Singleton EG, Pedarriosse AM, Lancrenon S, Rames A, Aubin HJ, et al. The Modified Reasons for Smoking Scale: factorial structure, gender effects and relationship with nicotine dependence and smoking cessation in French smokers. Addiction. 2003;98(11):1575-83.

10. Piper ME, Piasecki TM, Federman EB, Bolt DM, Smith SS, Fiore MC, et al. A multiple motives approach to tobacco dependence: the Wisconsin Inventory of Smoking Dependence Motives (WISDM-68). J Consult Clin Psychol. 2004;72(2):139-54.

11. Shenassa ED, Graham AL, Burdzovic JA, Buka SL. Psychometric properties of the Wisconsin Inventory of Smoking Dependence Motives (WISDM-68): a replication and extension. Nicotine Tob Res. 2009;11(8):1002-10.

12. Souza ES, Crippa JA, Pasian SR, Martinez JA. Factorial structure of the Brazilian version of the Modified Reasons for Smoking Scale [Article in Portuguese]. Rev Assoc Med Bras. 2009;55(5):557-62.

13. de Souza ES, Crippa JA, Pasian SR, Martinez JA. Modified Reasons for Smoking Scale: translation to Portuguese, cross-cultural adaptation for use in Brazil and evaluation of test-retest reliability. J Bras Pneumol. 2009;35(7):683-9.

14. Floyd FJ, Widaman KF. Factor analysis in the development and refinement of clinical assessment instruments. Psychol Assess. 1995;7(3):286-99.

15. Heatherton TF, Kozlowski LT, Frecker RC, Fagerström KO. The Fagerström Test for Nicotine Dependence: a revision of the Fagerström Tolerance Questionnaire. Br J Addict. 1991;86(9):1119-27.

16. Kaiser HF, Rice J. Little jiffy, mark IV. Educ Psychol Meas. 1974;34:111-17.

17. Cortina JM. What is coefficient alpha? An examination of theory and applications. J Appl Psychol. 1993;78(1):98 104.

18. Cichetti DV. A new measure of agreement between rank ordered variables. Proc Am Psychol Assoc. 1972;7:17-8.

19. Fleiss JL, Cohen J. The equivalence of weighted kappa and the intraclass correlation coefficient as measures of reliability. Educ Psychol Meas. 1973;33:613-19.

20. Anderson TW. An introduction to multivariate statistical analysis. New York: Wiley; 1985.

21. Costa PT Jr, McCrae RR, Bosse R. Smoking motive factors: a review and replication. Int J Addict. 1980;15(4):537 49.

22. Tiffany ST. A cognitive model of drug urges and drug use behavior: role of automatic and non-automatic processes. Psychol Rev. 1990;97(2):147-68.

23. Mackay J, Amos A. Women and tobacco. Respirology. 2003;8(2):123-30.

24. Perkins KA. Sex differences in nicotine versus non-nicotine reinforcement as determinants of tobacco smoking. Exp Clin Psychopharmacol. 1996;4:166-77.

25. Perkins KA. Smoking cessation in women. Special considerations. CNS Drugs. 2001;15(5):391-411.

26. Cawley J, Markowitz S, Tauras J. Lighting up and slimming down: the effects of body weight and cigarette prices on adolescent smoking initiation. J Health Econ. 2004;23(2):293-311.

27. Honjo K, Siegel M. Perceived importance of being thin and smoking initiation among young girls. Tob Control. 2003;12(3):289-95.





* Study carried out at the University of São Paulo at Ribeirão Preto School of Medicine and the University of São Paulo at Ribeirão Preto School of Philosophy, Science and Language, Ribeirão Preto, Brazil.
Correspondence to: José Antônio Baddini Martinez. Rua Salgueiro, 130, CEP 14040-210, Ribeirão Preto, SP, Brazil.
Tel 55 16 602-2531. E-mail: jabmarti@fmrp.usp.br
Financial support: None.
Submitted: 8 May 2010. Accepted, after review: 8 July 2010.
** A versão completa em português deste artigo está disponível em www.jornaldepneumologia.com.br




Sobre os autores

Elisa Sebba Tosta de Souza
Graduate Student in Clinical Medicine. Department of Clinical Medicine, University of São Paulo at Ribeirão Preto School of Medicine, Ribeirão Preto, Brazil.

José Alexandre de Souza Crippa
Professor. Department of Neurosciences and Behavioral Sciences, Division of Psychiatry, University of São Paulo at Ribeirão Preto School of Medicine, Ribeirão Preto, Brazil.

Sonia Regina Pasian
Professor. Department of Psychology and Education, University of São Paulo at Ribeirão Preto School of Philosophy, Science and Language, Ribeirão Preto, Brazil.

José Antônio Baddini Martinez
Associate Professor. Department of Clinical Medicine, University of São Paulo at Ribeirão Preto School of Medicine, Ribeirão Preto, Brazil.