DSM-IV and the South Oaks Gambling Screen: diagnosing and assessing pathological gambling in Turkey

DSM-IV and the South Oaks Gambling Screen:Diagnosing and Assessing Pathological Gambling

in Turkey

Ibrahim DuvarciAzmi Varan

Hakan CoskunolMehmet A. Ersoy

Ege University Medical School Psychiatry Department

The main purpose of this study was to investigate the effectiveness of the DSM-IV diag-nostic criteria and the South Oaks Gambling Screen (SOGS) in identifying Turkishpathological gamblers. Fifty-nine subjects participated in the study. The subjects werediagnosed as either pathological gamblers or not (comparison group) through the useof the DSM-IV criteria and were given the Turkish version of the SOGS. Four of the tenDSM-IV criteria were found to be problematic in the diagnosis of Turkish pathologicalgamblers. The data concerning reliability and validity of the Turkish version of theSOGS suggested that the SOGS can be used as a reliable and valid instrument in identi-fying Turkish pathological gamblers. Most (16 out of 20) of the items of the SOGSappear to work well in discriminating pathological gamblers from the subjects in thecomparison group. In the case of the two DSM-IV criteria and the four SOGS itemsthat failed to discriminate, cultural factors seemed to be responsible for the failure.

INTRODUCTION

Although the personal and social damage caused by gambling hasbeen known for centuries, pathological gambling has been officiallyrecognised as a diagnosable mental disorder only since 1980. The

Address correspondence to Ibrahim Duvarci, MD, Ege Universitesi, Tip Fakultesis, PsikiyatriAnabilim Dali, 35100-Bornova, Izmir, Turkey.

Journal of Gambling Studies Vol. 13(3), Fall 1997© 1997 Human Sciences Press, Inc. 193

JOURNAL OF GAMBLING STUDIES

third edition of Diagnostic and Statistical Manual of Mental Disorders(DSM-III) viewed pathological gambling as an "impulse control prob-lem", characterised by an irresistible impulse to gamble which leads todisruption and impairment of personal, social and occupational rolesand activities (American Psychiatric Association, 1980).

The scientific recognition of pathological gambling as a separateillness has started a series of studies attempting to understand the na-ture of the phenomenon. One important line of development regard-ing conceptualisation of pathological gambling has been an increasingawareness of similarities between addictive disorders and pathologicalgambling. Although pathological gambling has been listed under im-pulse-control disorders in the recently published DSM-IV (AmericanPsychiatric Association, 1994), there is a growing tendency to view thephenomenon as an "addiction" rather than an impulse-control prob-lem (Lesieur, 1988; Rosenthal & Lesieur, 1995).

The recognition of pathological gambling as a mental disorderand the subsequent efforts to understand its nature is no doubt closelyrelated to the fact that gambling is emerging as a major mental healthproblem (Volberg, 1994). Legalisation policies and increasing socialacceptance of gambling have led to an increase both in the number ofpeople who gamble and in expenditure on gambling (Ladouceur &Mayrand, 1987; Walker, 1992).

Situated between Asia and Europe, Turkey, for the last 75 years,has been trying to establish a modern, secular, westernised state on theruins of the Ottoman Empire that had its roots in eastern traditionsand teachings of Islam. Although gambling is strictly forbidden by Is-lam and prohibited in public places by the Turkish legislation, duringthe last decade Turkey has witnessed a dramatic increase both in legal-isation policies and in popularity of gambling activities.

Traditionally, illegal gambling in Turkey has been housed in "kah-vehane"s and "kulup"s. Kahvehanes, which literally means "coffeehouse", are places where Turkish men spend their leisure time drink-ing tea or coffee and playing various games such as cards, backgam-mon and dominoes. Although gambling in kahvehanes has alwaysbeen a very common activity, the amounts gambled are typically low.This is mainly because the kahvehanes are mostly used by people com-ing from low socio-economic status. Kulups, on the other hand, areplaces that provide gambling facilities to people from relatively highersocio-economic backgrounds, usually under a disguise of a social soci-ety or an association that requires membership. The amounts gambled

194

IBRAHIM DUVARCI, ET AL.

in kulups are typically higher compared to amounts gambled inkahvehanes.

Although gambling in public places is prohibited by law, Turkishgovernments have both sponsored and encouraged gambling since1926. For sixty years, between 1926 and 1986, the state-sponsored gam-bling activities in Turkey had been limited to only three games (Na-tional Lottery, Football Lotto and Horse Racing). During the last de-cade, however, four new state-sponsored gambling activities (two newtypes of Football Lotto, Scratch-and-Win and Lotto-6/49) have beenintroduced to the public. What seems to be responsible for this in-crease is the governments' attempts to find new sources of finance tocope with the increasing budget deficits.

The financial difficulties that have encouraged the state to intro-duce new types of gambling have also channeled millions of people intogambling. Faced with an annual inflation rate of 80-100% and ex-tremely low wages, the awards put forward by the lotteries and lotto'shave been seen by millions as "the only hope" to end their financialhardships. The state-sponsored lotteries and lottos, however, are notviewed by the public as gambling and the cases in which buying lotterytickets or playing lottos become a compulsive behaviour with accom-panying financial and psychological costs are relatively rare.

One other important line of development regarding increasingpopularity and legitimacy of gambling in Turkey is related to die legal-isation of gambling activities in the private sector. In 1983 the statelegalised private sector gambling by passing a law that allowed five-starhotels to open casinos. Since the state's purpose was to increaseTurkey's income from tourism, the casinos were initially allowed toprovide gambling services only to foreign tourists visiting Turkey. How-ever, the glittering world of the gambling machines was soon discov-ered by Turkish citizens who were illegally allowed to gamble by man-agement of the casinos. It was almost impossible for the casinos andthe state which taxed them to ignore the millions of dollars left at thegambling tables and machines by the Turks. In 1991 a new law permit-ting entrance of Turkish citizens into the casinos was passed.

What brought gambling to the attention of the public during thelast decade has been the increasing number of people gambling in thecasinos with large amounts of money. The amount of money gambledin the casinos in one year is estimated as one billion dollars (Akbulut,1997, personal communication). Moreover, the casinos situated at lux-urious hotels have opened the doors to gambling by women.

195


One noticeable consequence of the recent developments regard-ing gambling in Turkey has been the increase in the number of peo-ple seeking help for their gambling problems. Whereas the number ofpeople referred to our outpatient and inpatient clinics for gamblingproblems in one year did not exceed 10 in the early 80s, more than 40patients were diagnosed as pathological gamblers in 1996. Despite thefact that the Turkish mental health professionals have been faced withincreasing numbers of patients seeking help for their gambling prob-lems, studies on gambling in Turkey are almost non-existent.

The main purpose of this study was to investigate the effectivenessof the DSM-IV diagnostic criteria and the South Oaks GamblingScreen (SOGS) in identifying Turkish pathological gamblers. Follow-ing the publication of DSM-III in 1980, the DSM system has found agrowing interest among the Turkish mental health professionals and,at present, the DSM criteria are being quite widely employed in Tur-key for both clinical and research purposes. Despite its wide use,studies questioning the effectiveness of the DSM criteria in diagnosingmental disorders in Turkey are almost non-existent. Since gambling isinterwoven with many socio-cultural factors, we first decided to investi-gate the efficacy of the DSM criteria for pathological gambling in thediagnosis of Turkish pathological gamblers. Investigation of the SouthOaks Gambling Screen (SOGS) was undertaken to provide Turkishresearchers with an valid and reliable instrument that could be used toidentify pathological gamblers and encourage new research on gam-bling. The SOGS is a 20 item scale developed by Lesieur & Blume(1987) for identifying pathological gamblers and has been translatedinto many languages including French, German, Dutch, Spanish, Ital-ian, Cambodian and Lao (Lesieur & Blume, 1993).

METHOD

The study was carried out in the Psychiatry Department of theEge University in Izmir, Turkey, between October 1995 and July 1996.

Subjects

Fifty-nine subjects participated in the study. The subjects were ei-ther referred to the senior author for gambling problems or brought

196


to the attention of the authors by a number of sources (such as hospi-tal staff or friends) as people who could have gambling problems.Thirty-six (61 %) of these subjects were psychiatric inpatients or outpa-tients and eight (13.5 %) were hospital staff/employees or internshipstudents. The other 15 subjects were obtained from sources outsidethe hospital. Ten (17 %) were volunteer subjects interviewed inkahvehanes and cockfight settings and five (8.5 %) were volunteersubjects from the authors' neighbourhoods.

All subjects were interviewed by the senior author on the basis ofDSM-IV criteria for pathological gambling. The subjects were givenone point for each criterion they met, adding up to a total DSM-IVscore. Using the cut-off point of 5, at the end of the diagnostic inter-views, 35 (59.3 %) subjects were diagnosed as "pathological gamblers".The remaining 24 (40.7 %) subjects, who were not diagnosed aspathological gamblers, acted as the comparison group in the study.Some of the descriptive characteristics of each group are briefly sum-marised below.

Pathological Gamblers. Twenty-nine (82 %) of the 35 pathologicalgamblers were psychiatric inpatients or outpatients. In 14 (48.3 %) ofthese 29 subjects the main complaint was gambling problems whereas15 (51.7 %) were later identified as pathological gamblers while beingtreated for problems such as depression, alcoholism and panic attacks.The remaining six cases (18 %) of pathological gamblers consisted ofone hospital employee and five subjects obtained outside the hospitalsetting.

Thirty-two (91.4 %) of the pathological gamblers were male and 3(8.6 %) were female. The mean age was 38.37 (S.D.= 7.12, min-max= 24-51 years). More than half (57.1 %) had senior high schoolor university education, 14.3 % were junior high school graduates and28.6 % had only elementary school education. With regard to monthlyincome, almost seventy (68.6 %) percent had low income while 31.4 %had middle or high income. About two-thirds (62.9 %) of the subjectsin pathological gambling group were married, 11.4 % were single and25.7 % were separated, divorced or widowed.

Comparison Group. Seven (29 %) of the 24 subjects in the com-parison group were psychiatric inpatients or outpatients, 7 (29 %)were hospital staff/employees or internship students and the remain-

197


ing 10 (42 %) were subjects interviewed outside the hospital setting.The comparison group consisted of 20 (83.3 %) male and 4(16.7 %) female subjects. The mean age for the group was 37(S.D.= 8.58, min-max= 24-52 years). Half (50 %) of the group hadsenior high school education or more, 12.5 % were junior highschool graduates and 37.5 % had elementary school education. Morethan eighty percent (83.3 %) had low income while 16.7 % had mid-dle or high income. Almost eighty percent (79.2 %) of the subjectsin the comparison group were married, 8.3 % were single and12.5 % were separated, divorced or widowed.

Statistical analyses revealed no significant demographic differ-ences between pathological gamblers and the subjects in the compari-son group with respect to gender (Fisher's exact test, p> .05), age (t=.64, p> .05), education (x2 = .658, p> .05), income level (X

2 = 1.636,p> .05) and marital status (x2 = 1.901, p> .05).

Procedure

Following the diagnostic interview, each subject was given theTurkish version of SOGS. The SOGS was first translated into Turkishby three psychiatry residents and one clinical psychologist. The initialtranslated version of the SOGS was given to a group of 30 subjectsconsisting of psychiatric patients, nurses and internship students. Af-ter completing the test, each subject was interviewed about the clar-ity and comprehensibility of the items. In the case of inconsistentresponses on the test, the authors tried to find out whether this wascaused by the subjects' misunderstanding of the items due to proba-ble problems in translation or wording of items. Following the inter-views, each translated item was re-evaluated and revised if necessaryin the light of the data gathered from the interviews. The final Turk-ish version of the SOGS was almost a direct translation of the origi-nal instrument with respect to content and form. The only excep-tions were the first (types of gambling) and second (amount ofmoney gambled) items that had to be changed to adapt the items tothe Turkish culture (in line with the original authors' suggestions)(Lesieur & Blume, 1993).

198


RESULTS

Scores on DSM-IV for pathological gamblers and the subjects inthe comparison group are presented in Table 1.

As could be seen from Table 1, 80 % of the pathological gam-blers scored between 5-7 while only one out of 35 subjects scoredabove 8 points. In the comparison group about two-thirds (62.5 %)of the subjects had scores of either zero or one while 16.7 % scoredjust below the cut-off point of 5. The mean score on DSM-IV was6.43 (S.D.= 1.09, min-max== 5-9) for the pathological gamblers and1.38 (S.D.= 1.47, min-max= 0-4) for the subjects in the compari-son group.

Each DSM-IV criterion for pathological gambling was analysedseparately by Chi Square. All but the two (criteria 5 and 8) DSM-IVcriteria were found to be significantly related to the diagnosis ofpathological gambling. These data are presented in Table 2.

As it could be seen from data presented in Table 2, criteria 1, 2,3, 6, 7 and 9 seem to work better in discriminating pathological gam-blers from the subjects in the comparison group as compared to cri-

Table 1Scores on DSM-IV for Pathological Gamblers and the

Comparison Group

Score on DSM-IV

0123456789

PathologicalGamblers

n= 35 (%)

7 (20 %)14 (40 %)7 (20 %)6 (17.1 %)1 (2.9 %)

ComparisonGroup

n= 24 (%)

9 (37.5 %)6 (25 %)4 (16.7 %)1 (4.1 %)4 (16.7 %)

199

200 JOURNAL OF GAMBLING STUDIES

Table 2Item Analysis of DSM-IV Criteria for Pathological Gamblers and

the Comparison Group (Frequency (f) and Percent (%) of"positive" Responses)

DSM-IVPathologicalGambling Criteria

criterion 1criterion 2criterion 3criterion 4criterion 5criterion 6criterion 7criterion 8criterion 9criterion 10

PathologicalGamblers

f

353430141130323

288

n= 35

(%)

(100 %)(97.1 %)(85.7 %)(40 %)

(31.4 %)(85.7 %)(91.4 %)(8.6 %)(80 %)

(22.9 %)

ComparisonGroup

f

13551323010

n= 24

(%)

(54.2 %)(20.8 %)(20.8 %)(4.2 %)

(12.5 %)(8.3 %)

(12.5 %)

(4.2 %)

x2

19.71736.99924.8369.6412.818

34.34636.755

2.16732.7596.346

P

.00001*

.00001

.00001

.0019

.093

.00001

.00001

.263*

.00001

.0164*

* Fisher's exact test, two-tailed

teria 4 and 10. Criteria 5 and 8 failed to discriminate between thetwo groups.

The mean SOGS score was 11.31 (S.D.= 2.96; min-max= 6-17)for the pathological gamblers and 3.29 (S.D. = 3.52; min-max= 0-13) for the subjects in the comparison group. The difference be-tween the two groups was significant (t= 9.16, df= 43.74, p< .001).

The reliability of the Turkish version of SOGS was investigated bythe test-retest method. For the retest study 38 subjects who could bereached (18 pathological gamblers and 20 subjects from comparisongroup) were given the test a second time 30 or more days after the firstadministration. Test-retest correlation was found to be r= .95.

The criterion validity of the Turkish version of SOGS was investi-gated by cross-tabulating the patients' DSM-IV diagnoses of patholog-ical gambling with the same diagnoses based on the SOGS scores. Todo this, first of all a cut-off point for the SOGS scores had to be cho-


sen. To determine the cut-off point that would best discriminatebetween pathological gamblers and the subjects in the comparisongroup, different cutting points were examined in the light of the falsepositive and false negative percentages they yielded. These data arepresented in Table 3.

As it could be seen in Table 3, the cut-off point that yielded thelowest false negative and false positive percentages (and thus has thehighest sensitivity and specificity rates) was the score of 8. Using thescore of 8 as the cut-off point, the patients were classified as eitherpathological gamblers or not. To cross-check the validity of the SOGS,agreement between DSM-IV and the SOGS diagnoses of pathologicalgambling was investigated. The relevant data is presented in Table 4.

As it can be seen from the data presented in Table 4, the Turkishversion of SOGS and DSM-IV are significantly related in the diagnosisof pathological gambling (kappa= .789).

Finally, each item of the Turkish version of SOGS was analysedindependently by Chi Square analyses. The results of the item analysisare presented in Table 5.

All but the four items (Item 16b, 16f, 16g and 16i) of the SouthOaks Gambling Screen were found to be significantly related to patho-logical gambling.

Table 3Cut-off Points for the Turkish Version of South Oaks Gambling

Screen and False Positive, False Negative, Sensitivity andSpecificity Rates They Yield

Cut-offPoint onSOGS

5678910

False Positivef (%)

8 (33.3%)7 (29.2%)5 (20.8%)3 (12.5%)1 (4 .2%)1 (4 .2%)

False Negativef (%)

001 (2 .9%)3 (8.6%)8 (22.9 %)

11 (31.4%)

Sensitivity

11

.971

.914

.771

.686

Specificity

.666

.708

.791

.875

.958

.958

201

DISCUSSION

The data presented in Table 1 and the mean score of 6.43 onDSM-IV suggests that the majority of Turkish pathological gamblersmeet 5-7 of the 10 criteria put forward by the DSM-IV.

The data presented in Table 2 point to several problems in theemployment of DSM-IV criteria for diagnosing pathological gamblingin Turkish samples. The 8th criterion (has committed illegal acts) doesnot discriminate between pathological gamblers and the subjects inthe comparison group. As could be seen in Table 2, only three subjectsin the pathological gambling group met the 8th criterion. These threecases were being tried for embezzlement at the time of study and nonewas yet convicted. The low percentage of illegal acts among the Turk-ish pathological gamblers (8.6 %) and the subjects of the comparisongroup (0 %) can be partly explained by the differences in the justicesystems. Most of the pathological gamblers in our study have reportedborrowing money from different sources such as banks, friends andrelatives without paying it back. Although it is considered as a judicialfact, failing to repay borrowed money or a loan is not regarded as an"illegal act" in the same sense that are fraud, embezzlement or forgery.A person who does not repay borrowed money or a loan faces distraintand, only if the assets the person has cannot meet the demands of thedistraint, is the person liable to be sentenced to prison. Moreover, inmany cases, the extended family system and the traditionally close tiesamong the family members, lead the family to own and repay the gam-


Table 4Agreement of DSM-IV Diagnoses with the Turkish Version of

South Oaks Gambling Screen Diagnoses of Pathological Gambling

Diagnosis by DSM-IV Criteria

Pathological Gambler n = 35Comparison Subject n = 24

Diagnosis by the SOGS(cut-off point 8)

PathologicalGambler

32 (91.4 %)3 (12.5 %)

ComparisonSubject

3 (8.6 %)21 (87.5 %)

IBRAHIM DUVARCI, ET AL. 203

Table 5Item Analysis of South Oaks Gambling Screen for Pathological

Gamblers and the Comparison Group (Frequency (f ) and Percent(%) of "yes" Responses)

SOGSItems

456789101113141516a16b16c16d16e16f16g16h16i

PathologicalGamblersn= 35

f (%)

21 (60 %)19 (50.3%)32 (91.4 %)31 (88.6 %)31 (88.6 %)35 (100 %)27 (77.1 %)31 (88.6 %)22 (62.9 %)16 (45.7 %)24 (68.6 %)20 (57.1 %)10 (28.6 %)20 (57.1 %)18 (51.4 %)17 (48.6 %)8 (22.9 %)0

14 (40 %)0

ComparisonGroupn= 24

f (%)

2 (8.3 %)4 (16.7 %)6 (25 %)

11 (45.8 %)10 (41.7 %)14 (58.3 %)2 (8.3 %)5 (20.8 %)5 (20.8 %)1 (4.2 %)5 (20.8 %)3 (12.5 %)3 (12.5 %)3 (12.5 %)03 (12.5 %)1 (4.2 %)1 (4.2 %)00

x2

15.9978.470

27.40512.67814.77417.55926.97127.46410.12911.98112.98211.9292.140

11.92917.7618.2673.8471.483

12.586

P

.00006

.00361

.00001

.00037

.00012

.00003*

.00001

.00001

.00146

.00054

.00031

.00055

.143

.00055

.00003

.0040

.069*

.40*

.00039

*Fisher's Exact Test, two-tailed

bling member's debts and save the family the "humiliating" experi-ence of the distraint. Thus, even if the gambler lies and misinformsthe bank in taking the loan and fails to repay, in the eyes of the lawhe/she is no different than a decent citizen who has borrowed moneyfrom the bank and has failed to pay it back because of financial diffi-culties.


Another problem in using the DSM-IV criteria for diagnosingTurkish pathological gamblers is related to the 5th criterion (gamblingas a way of escaping from problems or of relieving dysphoric mood).In most of our pathological gambling cases gambling appeared as the"cause" of the problems rather than an escape response given to thealready existing problems. This statement, however, should be treatedcautiously since it may be that the pathological gamblers had "noproblems" prior to the gambling problem. Data gathered during thediagnostic interviews suggest that these subjects did indeed have prob-lems before they started to gamble heavily, however these problemsmost of the time were not regarded as the "causes" of the pathologicalgambling neither by the subjects nor by the authors. Moreover, if aclinician or a researcher looks for problems in a gambler's life he/shewill certainly come up with many but this would not necessarily meanthat the problematic gambling behaviour is an inevitable escape re-sponse to these problems. In our view, "gambling as a way of escapingfrom problems" is a problematic conceptualisation and could be leftout as it is with substance dependency.

As could be seen from Table 2, the 10th and the 4th criteria signifi-cantly discriminate between pathological gamblers and the subjects inthe comparison group. These two criteria, however, seem to be lesseffective than the remaining six criteria as indicated by the x2 valuespresented in Table 2. This might have stemmed from "construing" or"interpreting" the DSM criteria. Would it, for example, suffice for thesubject to report subjective feelings of restlessness or irritability, orshould restlessness or irritability be observable by others to meet the 4th

criterion? Would only one bailout (10th criterion) be enough orshould the interviewer look for repeated acts of bailout? If the inter-viewer decides on the latter, how many bailouts should he/she look forto decide that the subject meets the 10th criterion?

As mentioned in the Method section, the subjects in this studywere diagnosed based on the ten DSM-IV criteria and 35 of the 59subjects were diagnosed as pathological gamblers. If the four problem-atic DSM-IV criteria discussed above (the 4th, 5th , 8th and the 10th

criteria) were not employed in the diagnosis process (i.e. if the diag-nosis was based on 6 of the 10 DSM-IV criteria) only 2 (5.7 %) of the35 pathological gamblers would drop beneath the cut-off line of 5 sug-gested by the DSM-IV, and would no longer be diagnosed as patholog-ical gamblers. In other words, exclusion of these four criteria has onlya slight effect on the diagnoses made even when the cut-off point of 5

204


was maintained. Although the sample of this study allows for conclu-sions that could at best be suggestive, this finding seems to call forcaution in employing DSM-IV diagnostic criteria in cultures outsidethe USA, especially in the case of disorders that are more prone tocultural influences such as pathological gambling.

Another point that should be made in relation to DSM-IV criteriafor pathological gambling is related to the 2nd criterion. In almost all(97.1 %) of our pathological gamblers, we have observed gamblingwith increased amounts of money. However, in only a few of thesecases the increase in the amount of money gambled was "for achievingthe desired effect". Most of the time, the reported increases in theamount of money gambled were attributed to chasing. Most of thepathological gamblers reported that as their loses mounted, they hadto gamble with larger amounts in order to regain in the shortest timepossible what they had lost.

The preliminary data concerning reliability and validity of the Turk-ish version of SOGS suggests that the SOGS can be used as a reliable andvalid instrument in identifying Turkish pathological gamblers. The useof DSM-IV based diagnosis for the criterion validity of the SOGS mayappear problematic in the light of the four DSM-IV criteria that do notseem to work as well as the others. However, maintaining the cut-offpoint of 5 for the DSM diagnosis of pathological gambling seems tojustify both the classification made and the use of this classification forthe criterion validity of the Turkish version of SOGS.

Most (16 out of 20) of the items of SOGS appear to work well indiscriminating pathological gamblers from the subjects in the compar-ison group. In the case of the four items that have failed to discrimi-nate, cultural factors seem to be responsible for the failure. In Turkey,the percentage of employed women is very low. Even when the wifeworks, her income, most of the time, is seized by the husband who hasthe authority within the traditionally patriarchal Turkish society. Thissystem seems to be especially true for the families of low socio-eco-nomic status. Since the sample of the present study consists mostly ofmales with low income, Item 16b (borrowing money from spouse)does not appear to be a critical factor in the identification of thepathological gamblers. Borrowing money from loan sharks (16f), onthe other hand, is viewed rather as the last resort even by the patholog-ical gamblers and is avoided unless the person feels really desperateand hopeless. Loan sharks ask for incredibly high interest rates oververy short periods of time. For this reason, borrowing from a loan

205

shark is not a preferred way of finding money and thus leads to nosignificant differences between pathological gamblers and the subjectsin the comparison group.

Finally, with regards to Items 16g and 16i, neither stocks, bondsand other securities nor using checks are common ways in the Turkishculture (especially among people from low socio-economic class) ascould be seen from the extremely low percentages presented in Table5. Thus, neither cashing in securities like stocks, bonds etc. nor bor-rowing on one's checking account (passing bad checks) seem to be adiscriminating factor in identifying of pathological gamblers.

The data concerning the employment of DSM-IV criteria in theidentification of Turkish pathological gamblers call for careful study ofthe DSM-IV criteria as these may apply to the Turkish population inlight of the cultural differences discussed above. The Turkish versionof SOGS, on the other hand, seems to appear as a reliable and validinstrument in identifying Turkish pathological gamblers. Future re-search may concentrate on the DSM-IV criteria and the SOGS itemsthat do not seem to work as well as expected with the Turkish subjectsand suggest ways to improve the efficiency of both of the methods inidentifying pathological gamblers in Turkey.

REFERENCES

Akbulut, E. (1997). The Chairman of "Talih Oyunlan Mahalleri Isletenler Dernegi" (The Associa-tion of Managers of Gambling Vicinities), Istanbul, Turkey.

American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders, ThirdEdition, Washington, D.C., (pp. 291-293).

American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders, FourthEdition, Washington, D.C., (pp. 615-618).

Ladouceur, R. & Mayrand, M. (1987). Depressive behaviors and gambling. Psychological Reports, 60,1019-1022.

Lesieur, H.R. & Blume, S.B. (1987). The south oaks gambling screen (SOGS): A new instrument forthe identification of pathological gamblers. Am J Psychiatry, 144:9, September, 1184-1188.

Lesieur, H.R. (1988). Altering the DSM-III criteria for pathological gambling. J of Gambling Behavior,Vol 4(1), Spring, 38-47.

Lesieur, H.R. & Blume, S.B. (1993). Revising the south oaks gambling screen in different settings.Jof Gambling Studies, Vol. 9(3), Fall, 213-223.

Rosenthal, R.J. & Lesieur, H.R. (1995). Pathological gambling and criminal behavior. Draft ofchapter in Louis B. Schlesinger (ed.). Exploration in Criminal Psychopathobgy: Clinical Syndromeswith Forensic Implication.

Volberg, R.A. (1994). The prevalence and demographics of pathological gamblers: Implications forPublic Health. Am J Public Health; 84, 237-241.

Walker, M.B. (1992). The psychology of gambling. International series in social psychology (Paperbackedition, 1995). Oxford, UK: Butterworth-Heinemann Ltd., (p. 93).

Received January 27, 1997; revised July 29, 1997; final revision December 2, 1997; accepted Decem-ber 23, 1997.


DSM-IV and the South Oaks Gambling Screen: diagnosing and assessing pathological gambling in Turkey

Documents