On the construct validity of the sensation seeking scale in a non-english-speaking culture

On the construct validity of the sensation seeking scale in a non-english-speaking culture

Person. individ. Dtx 7, No. 3. pp. 431434, in Great Britain. All rights reserved 0191-8869186 1986 Ltd On the construct validity of the Sensation ...

419KB Sizes 18 Downloads 57 Views

Person. individ. Dtx 7, No. 3. pp. 431434, in Great Britain. All rights reserved

0191-8869186

1986

Ltd

On the construct validity of the Sensation Seeking non-English-speaking culture MENUCHA

Tel-Aviv

University, (Received

Scale in a

BIRENBAUM*

Ramat-Aviv,

Tel-Ah,

Israel

I August 1985)

Summary-The purpose of the present study was to examine the structure validity of Zuckerman’s Sensation Seeking Scale in a non-English-speaking culture. A nonmetric multidimensional scaling technique (SSA) was employed for analyzing the responses of 272 Israeli male applicants for a job for which sensation seeking was of special relevance. The general findings replicated the structure postulated by Zuckerman. Only I in 8 of the items were mislocated, all of which were indicated as inappropriate in previous factor-analytic studies conducted in other nations.

INTRODUCTION

In reviewing the stages in the development of the Sensation Seeking Scale (SSS), since its first introduction in the early 1960s. Zuckerman (1979) stresses the fact that at first the construct was conceived of as a unidimensional one, but then empirical results cast light on the more complicated nature of the construct. Based on factor-analytic studies, the following four distinct subfactors emerged: Thrill and Adventure Seeking (TAS); Experience Seeking (ES); Disinhibition (Dis); and Boredom Susceptibility (BS). Regarding the last three subfactors Zuckerman states that he did not expect to find them “and would never have constructed these scales using a purely rational method of scale development” (p. 102). Yet, these findings were replicated cross-nationally with British samples of males and females (Zuckerman, Eysenck and Eysenck, 1978). Comparison of the factorial structure between the American and British samples yielded, according to Zuckerman and his “remarkable cross-national reliability in spite of the differences in populations, testing procedures and form associates, used” (p. 105). The construction of Form V of the SSS was based on the results of this cross-national comparison. The assumption underlying the development of the SSS was that people differ in their optimal levels of stimulation and arousal and these differences influence their choices of life activity. The four different aspects of stimulation seeking are: Thrill and Adventure Seeking (TAS), which refers to enjoyment of internal arousal produced by physical risk-taking which mainly involves activities of the socially accepted type; Experience Seeking (ES), which involves seeking of arousal through the mind and senses, mainly expressing an anti-establishment self-fulfillment pattern; Disinhibition (Dis), which reflects the more traditional pattern of nonconformity through rebellion against strict codes of acceptable social behavior; and Boredom Susceptibility (BS); which refers to aversion to repetitive experiences of any kind, involving extreme tension and restlessness under conditions where such constancy is unavoidable. Additional support for the multidimensional structure of the sensation seeking construct comes from empirical studies which investigated the relations between the SSS and external criteria. The four SSS subscales were found to be differentially related to personality traits [see, for example, Gorman (1970), Bone (1972), Jacobs (1975). Eyesenck and Eysenck (1977). Birenbaum and Montag (1985)], to psychopathological factors (see Eysenck and Zuckerman, 1978; Zuckerman and Link, 1968; Blackburn, 1969; Zuckerman, 1979; Montag and Birenbaum, 1985) and to vocational choices (see Levin and Brown, 1975; Musolino and Hershenson, 1977; Biersner and LaRocco, 1983). However, recent studies concerning the psychometric properties of the latest version of the SSS (Form V) failed to replicate clearly the structure postulated by Zuckerman. Ridgeway and Russell (1980) examined the psychometric properties of the SSS in a Canadian sample of 181 females and 155 males and found the internal consistency of the subscales to be relatively low. Their conclusion was that “Further scale development thus appears necessary before widespread use of the SSS as a measure of stimulation seeking can be justified” (p. 664). Ball, Farnhill and Wangeman (1983), who investigated the factorial invariance across sex of Form V of the SSS in a sample of 335 female and 365 male Australian Ss, found several items that had not loaded on the expected factors, others that had not loaded at a significant level and a few that loaded on more than one factor. They concluded that the ES items as identified in their study were “more diverse in factorial composition than might be expected from previous results” (p. 1158). Although the SSS has been translated into various languages and used in studies conducted in different societies (see Berkowitz, 1967; Magaro. Smith, Cionini and Velicogna. 1979) no reported evidence exists concerning the factorial structure of Form V of the SSS in a non-English-speaking culture. According to Zuckerman himself “the status of the factors in translated forms of the SSS is still an open question” (Zuckerman et a/., 1978, p. 145). Furthermore, most of the studies which investigated the underlying structure of the SSS employed factor-analytic techniques. The relative!y small portion of common variance shared among the items of the instrument as reflected in the low communahties and the instability of the loadings on the expected factors (see Ball et al., 1983. p. 1157) raise doubt as to the applicability of factor-analytic techniques to this kind of data. One alternative method is nonmetric multidimensional scaling (MDS) which, like factor analysis, enables the exploration of the relations between the hypothesized structure and the results of the analysis, without imposing a particular structure. However, nonmetric MDS yields a more parsimonious solution than does factor analysis (Schlesinger and Guttman, 1969). Since it involves fewer parameters, the nonmetric MDS technique should yield a more stable solution.

*Present address: U.S.A.

CERL.

University

of Illinois at Urbana-Champaign,

431

252 ERL, 103 S. Mathews

Street, Urbana,

IL 61801,

432

NOTES AND SHORTER

COMMtiNICATIONS

Itern

Item

content

NO

TAS 3

Mountain

climbing

11

Dangerous

16

Water-skiing

17

Surf-board

20

Flying

an arrplane

21 23 28 38

Scuba

diving

“ding

Parachute

jumping

Divmg

off

highboard

Sailing

long

small

40

activities

distances

sailing

Skiing

fast

I” a

craft

down

a steep

slope

ES

2 *ES

4 6 9

I

. ‘2

DlS

Earthy

body

Explore

smells

a strange

crty

Marijuana

10

New

drugs

14 18

New

foods

19

Hippies

22 26 37

Clashing

Unplanned

trips

Homosexuals Dress

colors

in individual

way

Dis 1 I

26 ES

Wild

parties

12

Swingers

13

Get

25

Exciting

high

29

Exciting

30 32 33 35 36

Heavy

experiences dates drinking

Premarital Jet

sex

set

Sexy

movies

Alcohol

drinking

BS Getting

‘1.

18

2

Watching

5

Seeing

7

Predictable

people

8

Predictable

movies

\

.6 ES

TAS

20

lTAS

14 *ES

l%

. 40 TAS

\

l

with:

’ ES

TAS

.3

bored

17 ‘TAS 38 TAS

Fig. 1. Two-dimensional

\

movies the

15

Home-movies

24

Predictable

27

Staying

31

Being

same

several

times

faces

slides friends

home a bore

34

Insulting

39

Dull

others

people

SSA of the SSS items (N = 272)

The purpose of the present study was to investigate the underlying employing a nonmetric MDS technique for data analysis.

structure

of the SSS in a non-English-speaking

culture,

METHOD

The sample consisted of 272 Israeli male applicants for security-related jobs. The mean age of the Ss was 22.4 yr with a standard deviation of 1.6; their mean education was 12.5 yr of schooling; 96% of the Ss were born in Israel. Instruments

Form V of the SSS (Zuckerman, 1979) is a 40-item forced-choice questionnaire which yields four subscale scores and a Total score. The four subscales which consist of 10 items each are: TAS (Thrill and Adventure Seeking); ES (Experience Seeking); Dis (Disinhibition); and BS (Boredom Susceptibility). The properties of the Hebrew version of the questionnaire were studied by Birenbaum and Montag (1985). The findings regarding the predictive validity of the instrument were similar to those reported by Zuckerman and his associates (see Birenbaum and Montag, 1985; Montag and Birenbaum, 1985).

An MDS technique SSA-I (Smallest Space Analysis I) developed by Guttman and Lingoes (Lingoes, 1973, 1979; Lingoes and Guttman, 1979) was used for data analysis. SSA-I is a nonmetric computer program whose aim is to present graphically the relative size of the observed relationships between variables. The variables are presented as points in the derived geometrical space, the correlations serving as measures of distance. The level of measurement is taken as ordinal. The SSA technique is often applied to the correlation matrices either as an alternative to classical factor analysis or as a complementary procedure. For a detailed discussion about the relationship between SSA and factor analysis see Guttman (1954, 1958. 1982).

NOTES ANV SHORTER COMMUNICATIONS

433

RESULTS In order to test whether the pattern of intercorrelations among the 40 SSS items corresponds to the structure postulated by Zuckerman, the intercorrelation matrix was subjected to SSA-I. The goodness-of-fit for the SSS as measured by the coefficient of alienation (Lingoes, 1979) was 0.35 in two dimensions and reduced to 0.25 in three dimensions. However, since the structure was already evident in the two-dimensional case, only that solution will be given. Figure 1 presents the two-dimensional solution. Lines are drawn in the figure to partition the space into regions that represent the four SSS subfactors postulated by Zuckerman. As can be seen in the figure, 35 out of the 40 items are located in the expected regions. Only 1 in 8 of the items are mislocated: 3 of these items belong to the ES scale; 2 of them fell into the TAS region (No. 14 and No. 4) and 1 fell in the BS region (No. 26). The other 2 items originally belonged to the Dis scale, one of them (No. 28) fell into the ES region and the other (No. 32) into the BS region. Most of the TAS items cluster in a relatively small area, whereas the BS items are scattered over a large area. DISCUSSION

The pattern of intercorrelations among the SSS items as presented in the two-dimensional SSA map lends substantial support to Zuckerman’s contention regarding the underlying structure of the SSS. Most of the SSS items cluster in four distinct regions which correspond to the four subfactors (or facets) of the construct as postulated by Zuckerman: TAS (Thrill and Adventure Seeking); ES (Experience Seeking); Dis (Disinhibition); and BS (Boredom Susceptibility). Only a few items fell in regions other than the hypothesized ones. The good replicability of the TAS items and the weak one of the BS items are in line with previous findings (see Zuckerman rt al., 1978: Ball er al., 1983). The five mislocated items had already been pointed out in previous studies as problematic ones. Item No. 14, which refers to the desire to try new foods, is classified by Zuckerman as an ES item. In the current study, as well as in a study conducted on an Australian sample, this item fell among the TAS items. According to Zuckerman (1979) this item came from the Form IV BS scale (p. 106). Item No. 25, which refers to enjoyment of frightening and exciting experiences, belongs according to Zuckerman’s classification to the Dis scale. In the Australian study, as well as in the current one, this item fell among the ES items. Zuckerman (1979) claims that this item was shifted from the ES to the Dis scale in Form V of the SSS (p. 106). As a matter of fact, conceptually this item has more in common with ES items than with Dis items. Item No. 4, which refers to preference of earthy body smells, and No. 26, which expresses appreciation of clashing colors and irregular forms in painting, were classified by Zuckerman as ES items. In the present analysis the former item fell in the TAS region, whereas the latter fell in the BS region. In the Australian analysis both items failed to reach meaningful loadings on their expected factor. Zuckerman indicated that Item No. 4 is one of two items “that had not been in any of the previous factor scales” (p. 106). Item No. 32, which refers to premarital sex, belongs to the Dis scale according to Zuckerman’s classification. In the current analysis it was located in the BS (!) region whereas in the Australian study it fell with similar loadings on two factors: Dis and ES. It is worth noting that in the current study, as well as in previous ones [see, for example, Levin, Montag and Comrey (1983)], SSA yielded results which in general were similar to those of factor-analytic techniques. However, by mapping interpoint distances the SSA can provide additional information concerning the interlocation of the factors and the items within them [see, for example, Schlesinger and Guttman (1969) Levin, Karni and Frankel(l978) and Guttman and Shoham (1982)]. A close examination of item location along the vertical dimension indicates that with a few exceptions it refers to individual vs social activities. The lower part involves activities which are usually carried out on an individual basis such as diving off a highboard (No. 28) scuba diving (No. 21) parachute jumping (No. 23) flying an airplane (No. 20). The upper part refers to social activities such as partying (No. I). seeking pleasure around the world with the ‘jet set’ (No. 33). associating with ‘swingers’ (No. 12) etc. The horizontal dimension on the other hand refers to conventional vs nonconventional types of activities or preferences. The items to the right in each region refer to less conventional activities than do those to the left. For example, in the TAS region, sailing long distances in small sailing crafts (No. 38) vs flying an airplane (No. 20), in the ES region, trying new drugs (No. 10) vs meeting homosexual persons (No. 22). in the Dis region, getting high by using alcohol or drugs (No. 13) vs having a couple of drinks (No. 36), in the BS region, perceiving boredom as the 11’urst social sin (No. 31) vs getting bored with dull people (No. 39). Obviously, in the current version of the questionnaire these two dimensions are confounded with the original ones. A refinement of the instrument by adding new facets is needed in order to test these hypothesized dimensions in a controlled manner. REFERENCES

Ball 1. L., Farnill

D. and Wangeman J. (1983) Factorial invariance across sex of the Form V of the Sensation Seeking Scale. J. Person. sot. PsychoI. 45, 1 15&l 159. Berkowitz W. R. (1967) Use of the sensation-seeking scale with Thai subjects. Psychoi. Rep. 20, 635-641. Biersner R. J. and LaRocco J. M. (1983) Personality characteristics of U.S. Navy divers. J. occup. Psychol. 56, 329-334. Birenbaum M. and Montag I. (1985) Patterns of relationship between the 16PF and Zuckerman’s Sensation Seeking Scale. Multivar.

exp. clin. Res.

Blackburn R. (1969) Sensation seeking, impulsivity and psychopathic Bone R. N. (1972) PSI correlations with the MPI and Zuckerman’s Inventory Research and Development Notes (Edited by Lanyon R. Eysenck S. B. G. and Eysenck H. J. (1977) The place of impulsiveness Br. J. sot. chin. Psychoi.

personality. J. con.sulr. clin. P.sychoi. 33, 571-574. sensation seeking scale. In Psychological Screening I.), No. 3. in a dimensional system of personality description.

16, 57-68.

Eysenck S. B. G. and Zuckerman M. (1978) The relationship between sensation seeking and Eysenck’s dimensions of personality. Br. J. Psycho/. 69, 483487. Gorman B. S. (1970) 16PF correlates of sensation seeking Pq~chol. Rep. 26, 741 742. Guttman L. (1954) A new approach to factor analysis: the radex. In Ma/hema/ica/ Thinking in the Social Sciences (Edited by Lazarsfeld P. F.). pp. 258-348. Free Press, Glencoe, III. Guttman L. (1958) Order analysis of correlation matrices. In Handbook cf Mulrirariate E_~perimenfu/ P.Tycho/ogy (Edited by Cattell R. B.). pp. 438458. Rand-McNally, Chicago, Ill. Guttman L. (1982) Addendum; facet theory. smallest space analysis, and factor analysis. Percepl. Mar. Skills 54, 491493.

434

NOTES

AND

SHORTER

COMMUNICATIONS

Guttman R. and Shoham I. (1982) The structure of spatial ability items: a faceted analysis. Percept. Mot. Skill.7 54,487493. Jacobs K. W. (1975) 16PF correlates of sensation seeking: an expansion and validation. Psychol. Rep. 37, 1215~1218. Levin B. H. and Brown W. E. (1975) Susceptibility to boredom of jailers and law enforcement officers. Psychol. Rep. 36, 190.

Levin J., Karni

E. and Frankel

Y. (1978). Analysis

of the Tennessee

Self Concept

Scale as a faceted

instrument.

Psycho/.

Rep. 43, 619-623.

Levin J., Montag I. and Comrey A. L. (1983) Comparison of multitrait-multimethod, factor, and smallest space analysis on personality scale data. Psychol. Rep. 53, 591-596. Lingoes J. C. (1973) The Guttman-Lingoes Nonmetric Program Series. Mathesis Press, Ann Arbor, Mich. Lingoes J. C. (Ed.) (1979) Geometric Representations of Relational Data: Readings in Multidimensional Scaling, 2nd edn. Mathesis Press, Ann Arbor, Mich. Lingoes J. C. (Ed.) and Guttman L. (1979) Nonmetric factor analysis: a rank reducing alternative to linear factor analysis. In Geometric Representations CI/ Relational Data: Readings in Multidimensional Scaling, 2nd edn. Mathesis Press, Ann Arbor, Mich. Magaro P., Smith P., Cionini L. and Velicogna R. (1979) Sensation-seeking in Italy and the United States. J. sot. PsychoI. 109,

159-165.

Montag I. and Birenbaum M. (1985) Psychopathological factors and sensation seeking. J. Res. Person. Musolino R. F. and Hershenson D. B. (1977) A vocational sensation seeking in high and low risk-taking occupations. J. vocal. Behav. 10, 358-365. Ridgeway D. and Russell J. A. (1980) Reliability and validity of the sensation seeking scale: psychometric problems in Form V. J. consult. clin. Psychol. 48, 662-664. Schlesinger I. M. and Guttman L. (1969) Smallest space analysis of intelligence and achievement tests. Psychoi. Bull. 71, 95.-

100.

Zuckerman M. (1979) Sensation Seeking: Beyond the Optimal Level of Arousal. Erlbaum, Hillsdale, N.J. Zuckerman M. and Link K. (1968) Construct validity for the Sensation Seeking Scale. J. consult. clin. Psychol. 32, 42W26. Zuckerman M., Eysenck S. and Eysenck H. J. (1978) Sensation seeking in England and America: cross-cultural, age, and sex comparisons. J. consult. clin. Psychol. 46, 139-149.