Tobacco Documents Online

ielberger State-Trait Anxiety ijsc. Ue Poog, 1980, 15, .B. Defares. first and second grade children. Press. (With J. Papay). ate-Trait Curiosity Inventory. In Press. (With R.A. Peters and abits and the smoking behavior of I of Consultin and Clinical de Crane, and S. . Russe tted to: Journal of Personality .'3 795 Journal of Personality Assessment, In Press PERSONALITY A.,iD SMORING BEHAVIOR C. D. Spielberger and G. A. Jacobs University of South Florida RIJNNZYG BEAD: Personality and Smoking 95-077 0-82--51 0 r M 11

This study investigated the relaticnship bet~een selected personality measures and the initiation and maintenance of smoking behavior. The Eysenck Personality Questionnaire (EPQ), the State-Trait Personality Inventory (STPI) and a Smoking Behavior Questionnaire were administered to 955 students (603 females; 352 males). Smokers had significantly higher scores than Non-Smokers on the EPQ Extraversion, Neuroticism and Psychoticism scales, and lower scores on the Lie Scale. Female Smokers had higher STPI Trait Anxiety scores than Non-SmokeFs, whereas male Smokers had lower anxiety scores than :Ion-Smokars. Female Current Smokers had lower Neuroticism and Trait Anxiety scores than Occasional and Ex- Smokers, and Ex-Smokers of both sexes had higher scores on the STPI Trait Curiosity scale than Current and Occasional Smokers. It was con- cluded that the initiation and maintenance of smoking behavior are in- fluenced by different personality factors. that smoking was assoc these findings either c.eaningful conclusions Matarazzo and Mat pretation of the liter "The results, meager a presence of a slightly 'tense' individuais an -(-1965; -p.-3771- The studies publi evidence that smoking dencies (Brackenridge Spilken, 1971). Smoke Eysenck's Psychoticism & Gry11s, 1979), which (Eysenck & Eysenck, 19

Personality & Smoking ; Behavior ship between selected personality :e of smoking behavior. The :he State-Trait Personalitv :estionnaire were administered Smokers had significantly Extraversion, Neuroticism and :he Lie Scale. Female Smokers i Non-Smokers, whereas male Smokers. Female Current Smokers ores than Occasional and Ex- higher scores on the STPI asional Smokers. It was con- of smoking behavior are in- 797 Personality and Smoking Behavior Personality & Smoking 2 The diverse literature on smoking and personality was reviewed by Smith (1970), who observed that smoking was positively associated with the follow- iag personality traits: extraversion (12 of 15 studies); antisocial tenden- cies (17 of 19 studies); and impulsive behavior (6 of 8 studies). Smith concluded that smokers were more extraverted and have more antisocial ten- dencies than non-smokers, and that the evidence is "reasonably convincing" that smokers are also more impulsive. Although there was some evidence that smoking was associated with nearoticism add anxiety, Smith considered these findings either inconsistent or based on too few studies to draw meaningful conclusions. Matarazzo and Matarazzo have offered a somewhat different inter- pretation of the literature on smoking and personality. In their view: "The results, meager and poorly supported as they are, suggest the..... presence of a slightly higher number of 'extravert', and 'neurotic', and 'tense' individuals among the smokers as compared to the non-smokers" The studies published since Smith's (1970) review provide further evidence that smoking is associated with extraversion and antisocial ten- dencies (Brackenridge & B1och, 1972; Ranekar & Dolke, 1970; Jacobs & Spilken, 1971). Smokers also have higher scores than non-smokers on Eysenck's Psychoticism Scale (Jamison, 1978; Powell, 1977; Powell, Stewart, 6 Grylls, 1979), which is assumed to measure impulsive, antisocial behavior (Eysenck & Eysenck, 1975). In addition, there is growing evidence that smokers are more tense, neurotic and anxious than non-smokers as inferred from higher scores on Eysenck's veuroticism Scale (Brackenridge S Bloc.h,

798 Personality & Smoking 3 1972; Gupta, Sethi, & Gupta, 1975; Powell at al., 1979) and the Taylor (1953) Manifest Anxiety Scale (Houston & Schneider, 1973; Schneider S Houston, 1970), a measure of trait anxiety (Spielberger, 1966, 1972). There are two major difficulties in interpreting the research findings on smoking and personality. First, the definition of 'smokers' and 'non- smokers' has varied from study to study. A second problem is the failure to distinguish between factors that influence people to begin to smoke and that contribute to the maintenance of smoking behavior once it has been established. In evaluating the association between personality and the initiation of smoking, current smokers and ex-smokers should be classified as smokers. In esamining the effects of personality variables on the maintenance of smoking behavior, current smokers an~ ax-smokers must obviously be considered as separate groups. It may also be important to distinguish occasional smokers (e.g., people who smoke from time to time but not every day) from regular smokers, ex-smokers and non-smokers. The primary goal of the present study was to investigate relationships between extraversion, neuroticism and psychoticism as measured by the Eysenck Personality Questionnaire (EPQ) and the initiation and maintenance of smoking behavior for American college students. Although the scales developed by Eyseack and his associates have been widely used in smoking research in England, Australia and India, no studies could be found in which these scales were used to investigate relationships between person- ality and smoking with American subjects. A second goal of this study was to examine the association between smoking behavior and trait anxiety as measured by the State-Trait Personality Inventory (STPI, Spielberger, 1979). On the basis of score higher than nc Neuroticism and Psyc ported in a number o (see Eysenck, 1980), males and females. Sub ects The subjects we enrolled in introduc university; the medi. take part in the stu< The sample consisted enrolled in iatroduc. Test Instruments The E9senck Pere Eysenclc; 1975), the c developed over the pc 1952; Eysenck, 1959; items. The EPQ yielc Psychoticism (P), the It also includes a Li The State-Trait to measure anxiety, c craits (Spielberger, similar in format and

: a1., 1979) and the Taylor aeider, 1973; Schneider & Spielberger, 1966, 1972). :erpreting the research findings nition of 'smokers' and 'non- second problem is the failure a people to begin to smoke and g behavior once it has been between personality and the z-smokers should be classified sonality variables on the iers and ea-smckers must It may also be important ?le who smoke from time to , as-smokers and non-smokers. as to investigate relationships _icism as measured by the :he initiation and maintenance iencs. Although the scales been widely used in smoking studies could be found in -elationships between person- second goal of this study was :havior and trait anxiety as aorv (STPI, Spielberger, 799 Personality & Smoking 4 On the basis of previous research findings, smokers were expected to score higher than non-smokers in trait anxiety, and on the EPQ Extraversion, Neuroticism and Psychoticism scales. Since sex differences have been re- ported in a aumber of previous studies of personality and smoking behavior (see Eysenck, 1980), these relationships will be separately evaluated for males and females. Method Sub ects The subjects were 955 undergraduate students (603 females, 352 males) enrolled in introductory level psychQlogy courses at a large urban state university; the median age was 19 years. These students volunteered to take part in the study and received course credit for their participation. The sample consisted of more than 70 percent of the students who were enrolled in introductory psychology at the time the study was conducted. Test Instruments The Eysenck Personality Questionnaire (EPQ). The EPQ (Eysenck 6 Eysenck; 1975), the most recent form of a series of personality inventories developed over the past 30 years by Eysenck and his colleagues (Eysen dc, 1952; Eysenck, 1959; Eysenck & Eysenck, 1968), consists of 90 true-false items. The EPQ yields scares on Extraversion (E), Veuroticism (Y) and Psychoticism (P), the three dimensions of Eysenck's personality theory. It also includes a Lie (L) scale. The State-Trait Personality Inventory (STPI). The STPI was developed to measure anxiety, curiosity and anger as emotional states and personality traits (Spielberger, 1979). This 60-item self-report inventory is similar in format and conception to the Stare-Trait Anxiety Invencorv

800 (Spielberger, Gorsuch, & Lushene, 1970). The three 10-item STPI trait scales require the respondents to report the frequency that they have experienced the specified personality characteristics by rating themselves on the following four-point scale: 1) Almost never; 2) Sometimes; 3) Often; 4) Almost always. Persons with high STPI scores on a particular per- sonality trait are more prone to experience the emotional states associated with that trait than persons with low scores. Although the primary interest in the present study was in trait anxiety, the STPI Trait Curiosity and Trait Anger scales were also administered for exploratory purposes.l The Smoking Behavior Questionnaire (SBQ). The SBQ is a 50-item self- report questionnaire designed to elicit specific information about students' I smoking behavior and the smoking habits of their fami.lies? In constructing the SBQ, a number of questionnaires used to evaluate smoking habits in pre- Taylor, & Solo:an,, vious investigations were carefully revie•aed (Clausen, 1968; Horn, Courts,/ 1959; Ikard, Green & Horn, 1969; Leventhal & Avis, 1976), and some of the items from these instruments were adapted for the present study. laforma- tion about the construction of the SBQ and the relation becween student smoking behavior and family smoking habits is reported by Spielberger, Jacobs, Crane, and Russell (in press). - Procedure In order to encourage and facilitate participation in the study by a large proportion of the students enrolled in introductory psychology courses, the testing sessions were scheduled immediately after the classes from whhirbh the students were recruited. Groups of 20 to 100 students were tested over a 10-month period. At the beginning of each testing session, the students were informed that the purpose of the study was to "learn more about the feelings and attitudes of college students". They were also informed that feedba about the research would students then read and si cipation in the study was pleting several questioffi The test instruments (a) the STPI; (b) the SBQ with standard instruction cigarette smoking habits, Current Smokers, Occasion. instructions defined an "( cigarettes from time to t: defined as someone who ha: with cigarettes, but neve_ "- The percentages of at Smokers, Occasional Smoke: Table 1. Current Smokers, fied as Smokers in detera reported in Table 1. Inse: Personalitv and the Initiz In evaluating the ass the initiation of smoking,

Personality & Smoking 5 ihe three 10-item STPI trait ze frequency that they have - acteristics by rating themselves ~st never; 2) Sometiaes; 3) Often; acores on a particular per- a the emotional states associated as. Although the primary interest the STPI Trait Curiosity and for exploratory purposes.1 L(Z) . The SBQ is a 50-item self- acific information about students' their families? In constructing a evaluate smoking habits in pre- Taylor, 6 Solom ed (Clausen, 1968; Horn, Courts,/ & Avis, 1976), and some of the for the present study. Iaforma- d the relation between student is reported by Spielberger, participation in the study by a in introductory psychology .ed immediately after the classes 'roups of 20 to 100 students were inning of eac:z testing session, e of the study was to "learn :ollege students". They were W 801 Personality 6 Smoking 6 also informed that feedback sessions to provide additional information about the research would be echeduled near the end of the term. The students then read and signed a Consent Form, advising them that parti- cipation in the study was entirely voluntary, and would consist of com- pleting several questionnaires. The test instruments were administered in the following order: (a) the STPI; (b) the SBQ; (c) the EPQ. The STPI and the EPQ were given with standard instructions. On the SBQ, which inquired about past and present cigarette smoking habits, students were asked to report whether they were Current Smokers, Occasional Smokers, Ex-Smokers or Non-Smokers. The - instructions defined an "Occasionaj Smoker" as: "Someone who smokes cigarettes from time to time but not everyday." A'^Ion-Smoker" was defined as someone who had never smoked, or had only experimented briefly with cigarettes, but never became a regular or Occasional Smoker. Results The percentages of students who indicated that they were Current Smokers, Occasional Smokers, Ex-Smokers or Non-Smokers are reported in Table 1. Current Smokers, Occasional Smokers and Ex-Smokers were classi- fied as Smokers in determining the percentage of female and male "Smokers" reported in Table 1. Insert Table 1 about here Personality and the Initiation of Smoking Behavior In evaluating the association between the personality measures and the initiation of smoking, the scores of Smokers and Non-Smokers on the I 0

802 Personality & Smoking 7 EPQ and the STPI scales were compared. The means and standard deviations for each EPQ and STPI scale are reported in Table 2. These data were evaluated in separate 2 a 2 factorial analyses of variance in which Groups (Smokers vs. Non-Smokers) and Sex were the independent variables; the results of these analyses are also reported in Table 2. Insert Table 2 about here The significant Groups main effects for all four EPQ scales indicated that Smokers had higher scores than Non-Smokers on Extraversion, aeuro- ticism and Psychoticism, and lower Lie scale scores. The significant 1 Sex main effects for the EPQ Neuroticism, Psychoticism and Lie scales reflected the fact that the females scored higher than the males on the Neuroticism and Lie scales, and lower on Psychoticism. None of the Groups by Sex interactions were significant for any of the EPQ scales. While the differences on the EPQ scales were in the same direction for both sexes, these differences were larger in magnitude for the females, _,f except for Neuroticism on which comparable differences were found. In the separate analyses for each sex, the female Smokers scored significantly higher than Non-Smokers on Psychoticism and Eatraversion, and sianifi- cantly lower on the EPQ Lie scale. Male Smokers also had significantly lower Lie scale scores than Non-Smokers. For the STPI scales, the only statistically significant differences were the Sex main effect and the Groups by Sex interaction for traic anxiety. These findings reflected the fact that fema.Ies had higher anxiety scores than males, and that female Smokers had higher anxiety scores than Non-Smokers, whereas male Smokers had lower anxiety scores than Non-Smokers. No a, in the separate analysee Personality and the Naii The mean EPQ and S: and Ex-Smokers are repor ance of smoking behavioz 3 a 2 factorial analyses and Ex-Smokers) and Sex analyses are also report arate analyses of the dE There were no stati overall analyses of the the Neuroticism aad Psyc on N and lower on P that Occasional Smokers of bc Current and Ea-Smokers, separate analyses for fe scores than the Occasion In the analyses of significant Groups effec Fx-Smokers of both seaes and Occasional Smokers. Trait Anxiety, which iac this scale than males, e

)2 Personality & Smoking 7 The means and standard deviations i- in Table 2. These data were zalyses of variance in which x were the independent variables; :eported in Table 2. - jut here s for all four EPQ scales indicated -Smokers on Extraversion, Neuro- 3cale scores. The significant i, Psychoticism and Lie scales :ed higher than the males on the i Psychoticism. None of the Groups : any of the EPQ scales. :ales were in the same direction _arger in magnitude for the females, ile differences were found. In the ile Smokers scored significantly and Extraversion, and signifi- a Smokers also had significantly .stically significant differences by Sex interaction for trait `act that females had higher .le Smokers had higher anxiety .okers had lower anxiety scores 803 Personality & Smoking 8 than Non-Smokers. No significant differences were found for either sex in the separate analyses of the STPI scales. Personality and the Maintenar.ce of Smoking Behavior The mean EPQ and STPI scores of the female and male Current, Occasional and Ex-Smokers are reported in Table 3. Associations between the mainten- ance of smoking behavior and each personality measure were evaluated in 3 x 2 factorial analyses of variance in which Groups (Current, Occasional and Ex-Smokers) and Sex were the independent variables. Summaries of these analyses are also reported in Table 3, along with the findings in the sep- arate analyses of the data for female and male students. t Insert Table 3 about here There were no statistically significant Group differences in the overall analyses of the EPQ scores, but sex main effects were found for the Neuroticism and Psychoticism measures. The females scored higher on N and lower on P than males, as was noted in the preceding analyses. Occasional Smokers of both sexes also had lower Lie scores than the Current and Ex-Smokers, who were quite similar on this scale. In the separate analyses for females, the Current Smokers had lower Veuroticism scores than the Occasional and Ex-Smokers. In the analyses of the data for the STPI scales, the statistically significant Groups effect for Trait Curiosity reflected the finding that Hz-Smokers of both sexes had higher scores on this measure than Current and Occasional Smokers. A significant Sex main effect was also found for Trait Anxiety, which indicated that females had higher anxiety scores on this scale than males, as was noted in the preceding analyses. In the

separate analyses for females, the significant Groups effect for the STPI trait anxiety measure reflected the fact that Current Smokers had lower anxiety scores than Occasional and Ex-Smokers. - Discussion In the present study, Smokers scored higher than Non-Smokers on the EPQ Extraversion, Neuroticism and Psychoticism scales, and Non-Smokers had higher Lie scores than Smokers. While these findings were generally con- sistent with results previously reported by other iavestigators,(Bracken- ridge & Bloch, 1972; Gupta et al., 1975; Powell et al., 1979), this is the first study to obtain such relationships for an American sample. The differences observed between Smokers and Non-Smokers on the EPQ scales were in the same direction for both sexes, but these differences were larger in magnitude for the females. The findings in the preseat study that females scored higher than males on the Neuroticism and Lie scales, and lower on Psychoticism, were consistent with the results reported for British subjects in the EPQ Test Manual (Eysenck & Eysenck, 1975). In evaluating the association between the STPI scales and the initia- tion of smoking, it was found that female Smokers had higher anxiety scores than Non-Smokers, whereas male Smokers had lower anxiety scores than Noa- Smokers. The results for the females are consistent with findings reported " by several previous investigators that smokers scored higher in anxiety than non-smokers (Houston & Schneider, 1973; Schneider & Houston, 1970), and support Matarazzo and Matarazzo's (1965) conclusion that there are, "...a slightly higher number of...'neurotic', and 'tense' individuals among smokers as compared to the non-smokers" (p. 377). The findings in the present study of a :or females and a negative of taking the sex of the e relationship betweea.smoki In evaluating the ass of smoking behavior, no si Occasional and Ex-Smokers scores. However, in the s Smokers had lower scores o A similar trend was also f statistically significant, the differences were large Female Current Smokers als, than Occasional and Ex-Smo; Taken together, the v who are higher in neurotic: smoking in order to reduce female Current Smakers.sco: than Occasional and ErSmoi effective tension reducer : an alternative explanation higher in neuroticism and t longitudinal prospective st evaluate these alternative An interesting new fii of both sexes had higher s, Current and Occasional Ssno: