VALIDITY OF SELF-REPORTED CRIMINAL OFFENCES AND TRAFFIC VIOLATIONS IN SCREENING OF DRIVING-WHILE-INTOXICATED OFFENDERS

Transcription

1 Alcohol & Alcoholism Vol. 1,. 6, pp , 1996 VALIDITY OF SELF-REPORTED CRIMINAL OFFENCES AND TRAFFIC VIOLATIONS IN SCREENING OF DRIVING-WHILE-INTOXICATED OFFENDERS I. CHANG* and S. C. LAPHAM Substance Abuse Research Program, Institute for Health and Population Research, The Lovelace Institutes, 1650 University Blvd. Suite 0, Albuquerque, NM 8710, USA (Received 5 March 1996; in revised form 1 July 19%; accepted July 1996) Abstract Many jurisdictions in the USA, Canada and some European countries use diagnostic methods to assess substance abuse problems of driving-while-intoxicated (DWI) offenders, to address the concern that, during DWI screening, offenders may not give accurate information on their criminal history and traffic violations to avoid referral to treatment. This study was designed to validate selfreported data, to assess the need for DWI agencies to access court records, and to obtain an offence profile for this population. DWI offenders (n = 7, mostly first-time) were randomly selected from those who attended the Lovelace Comprehensive Screening Program (LCSP). The self-reported data were compared with records retrieved from the Metropolitan Court in Albuquerque, New Mexico, USA. Three-quarters of the offenders had had at least one offence or traffic violation before this DWI arrest. Sixty-five per cent of the offenders with court records underreported their records. The high percentage of false self-reporting for a primarily first-time offender population indicates the need to use court records to verify self-reported data For multiple offenders, who have a much higher rate of criminal offences and traffic violations, checking self-reported data against court records becomes more important In addition, a questionnaire based on offence profile information could be used to obtain a more complete history of those offences. INTRODUCTION Many jurisdictions in the United States of America and Canada use diagnostic assessment methods to assess substance abuse problems of driving-whileintoxicated (DWI) offenders for the purposes of referral to further assessment and/or treatment (Mayhew et al, 199). European countries such as Austria, Germany and Switzerland use formalized diagnostic instruments for all problem drivers including those arrested for DWI (Mayhew et al., 199). Examples of these DWI screening programmes are the Alcohol Safety Action Program in Alaska, USA the DWI Screening and Treatment Program in Arizona, USA and the Alberta IMPACT Program in Alberta, Canada (Popkin et al, 1988; Mayhew et al, 199). Operation of these programmes is influenced by the philosophy of the court, budgetary constraints, personnel Author to whom correspondence should be addressed. training capabilities and choice of test instruments. Although screening procedures vary significantly (Popkin et al, 1988, Mayhew et al, 199), screening programmes generally include testing, interviewing, and referral services. Those services can be provided by the court or by private DWI service agencies or locally certified alcohol and drug evaluation specialists. Some programmes also have a component to monitor client's compliance with the treatment. During testing, diagnostic instruments are administered to evaluate the alcohol and drug use of DWI offenders and other psychological and behavioural characteristics associated with recidivism risk (Lapham et al, 1995). Often specially trained personnel interview offenders to understand further the circumstances of the arrest, and any family, medical, personal, or legal problems that may indicate a need for treatment services. For those referred to treatment, progress and compliance with treatment procedures can be monitored. Information gathered by a screening Medical Council on Alcoholism Downloaded from on 19 August 018

2 58 I. CHANG and S. C. LAPHAM programme can include: client demographics, family and health problems, measures of blood alcohol content (BACs) at arrest, prior alcohol and drug use, criminal history, prior DWI arrests, and other traffic violations (Popkin et ai, 1988; Lapham et ai, 1995). While some screening programmes are administered through the courts, most are contracted to public or private agencies. In the latter case, most of the information gathered during screening is self-reported by the offenders rather than based on the court arrest records. Popkin (1989) estimated that 7% and 8% of the DWI screening facilities in rth Carolina, USA, did not have access to driving records and court criminal records, respectively. Therefore, the reliability of these self-reported responses has been a concern [National Institute on Alcohol Abuse and Alcoholism (NLAAA), 199], especially when clients face the legal consequences of being referred to treatment. The accuracy of self-reported alcohol consumption has been studied by several investigators. Watson et ai (198) and Fuller et al. (1988) reported that alcoholics underestimated their alcohol consumption when their reports were compared to informant reports. Other studies have reached the opposite conclusion (Polich, 198; Hesselbrock et ai, 198), finding no significant differences between self-reports by alcoholics and informants. In a review of this issue, Babor et ai (1987) suggested that since researchers must rely on self-report data, we should focus more efforts on improving existing instruments to make them as accurate as possible. Despite numerous studies that have been conducted on the validity of self-reported alcohol consumption, the accuracy of self-reported information regarding criminal arrests and moving violations in the context of DWI screening has not been reported. Donovan et al. (1985) reported on the basis of a -year follow-up study that the number of traffic violations and other arrests can be used to predict future DWI convictions. Wilson and Jonah (1985) also concluded that drivers, who admitted driving while impaired, reported more violations and crashes. Information about prior criminal arrests and traffic violations is therefore critical to assess alcohol-related problems and predict future recidivism. The present study was designed to assess the validity of this information when collected through self-reporting during a screening process. Self-reported records at the Lovelace Comprehensive Screening Program (LCSP) and records of the Bernalillo County Metropolitan Court in Albuquerque, New Mexico were compared. Record searches of other agency files (such as courts in the neighbouring counties, police departments, the Motor Vehicle Division, and the FBI) were not conducted. This study was designed to: (1) assess the validity of self-reported criminal records and traffic violations; () evaluate the need of DWI screening agencies to assess court records; () obtain an offence profile of the offenders; () make suggestions for DWI screening agencies to minimize bias and enhance accuracy. METHODS Lovelace Comprehensive Screening Program The study population consisted of a random sample primarily of first-time DWI offenders referred by the Bernalillo County Metropolitan Court to the Lovelace Comprehensive Screening Program (LCSP) to assess alcohol-related problems between April, 1989 and October, 199, before LCSP gained access to Bernalillo County's court records. The LCSP consisted of registration, testing, and interviews. Standardized evaluation instruments were used during testing to identify the presence, nature and intensity of alcohol- or drug-related problems. Clients were required to receive a structured personal interview with a master's-degree-level counsellor. The counsellor reviewed and discussed the test results with the client, determined the circumstances of the DWI arrest, asked about previous arrest history and the history of other violations, and obtained information about any legal, social, family or medical problems that might have resulted from, or caused, the client's alcohol or drug abuse. The counsellor determined the severity of the client's alcohol abuse and referred the client for treatment when indicated. Clients' self-reports and court records During the interview with the counsellor, clients were asked about their prior DWI arrests, criminal offences, and traffic violations in the last years. Specific questions were: (1) Have you had any previous arrests for DWI? If so, how many times? () Have you ever been arrested for any other Downloaded from on 19 August 018

3 SCREENING DWI OFFENDERS 585 criminal offence(s)?... possession of illegal drugs?... petty larceny?... domestic violence?... getting into fights?... burglary?... dealing drugs?... other? () Have you received any other moving traffic violation(s), other than this arrest, during the past years? If so, how many times? For reckless driving?... running a stop sign?... not wearing a seat belt?... running a red light?... speeding?... other? Responses were directly entered onto a computer file by the counsellor. The court's database, containing information on arrests and charges since 198, was accessed by a research associate using a referral number assigned by the court. In rare cases, when the referral number did not allow access, the original client folder was pulled and the client's name and/ or social security number was used. In that situation, of the name, social security number, and birthday, at least two out of three had to match for the records to be considered as belonging to the same person. The records were obtained and entered onto a computer file at LCSP. Information obtained from the court included criminal and motor vehicle codes, violation dates, and descriptions for all charges and arrests. Review of LCSP and court records A total of 00 DWI offenders who attended the screening programme at LCSP were randomly selected. Due to incomplete information on 6 of them, a final sample of 7 offenders was used to form the study population. The cut-off date for selection was October 199, when LCSP obtained direct computer access to the court database. Study participants' LCSP and court records were carefully reviewed by a research associate. Court records were recoded according to the questions listed above that were asked during the interview. Offenders' previous DWIs, criminal offences, and traffic violations were of primary interest. The research associate classified charges and arrests into categories of criminal offences and traffic violations. The New Mexico Criminal and Traffic Law Manual (199) was used when classifications were not immediately clear. Charges and arrests that occurred before the date of the interview were included, as were traffic violations that occurred within years before the date of the interview. Secondary charges, in which clients were held in contempt of court, including failure to appear in court, failure to register for DWI school, failure to comply with the first-time offender programme, etc., were not the focus of the study and therefore were not counted. Court records were also tabulated to indicate the most frequent criminal offences and traffic violations. Categories included previous DWIs, domestic violence, assault and battery, obstructing or evading an officer, disorderly conduct or public drunkenness, traffic moving violations, licence or registration violations, etc. RESULTS Study participants Of this study population, % were Hispanic; % were non-hispanic White; 10% were Mexican National; 10% were Native American; and % were from other ethnic or racial groups. Seventy-eight per cent of the population were male. Clients' ages were: % under 6 years of age, 6% between 6 and 0, and 1% over 0. Thirty-nine per cent of clients were single; 5% were married or living together as married; 6% were divorced or separated. Demographic characteristics of the study population were similar to those of the entire DWI population (Chang et ai, 1996). Offence profile of DWI offenders Court records showed 67 offences for the 7 offenders (Table 1). Licence or registration violations were the most frequent offences (6%), followed by insurance violations (0%) and moving violations (1%). Moving violations included reckless driving, running a stop sign, running a red light, speeding, failing to keep proper lookout, failing to drive within the traffic lane, passing a school bus, and others. Thirty-eight offenders had 5 previous DWIs. Other misdemeanours included resisting or obstructing an officer (% of all offences and 5% of all offenders), domestic violence (% of all offences and % of all offenders), disorderly conduct or public drunkenness (% of all offences and % of all offenders), assault or battery (% of all offences and % of all offenders), criminal trespassing (% of all offences and % of all offenders), etc. The mean number of offences for all participants was.. For those with previous DWIs, the mean number increased to 6.0. Downloaded from on 19 August 018

4 586 I. CHANG and S. C. LAPHAM Table 1. Offence profile of the DWI offenders ranked by number of occurrences (based on court records) Type of offence. of occurrences %. of offenders %* Licence/registration violations insurance/proof of financial responsibility Moving violations! Previous DWIs Resisting/obstructing/refusing to obey an officer Equipment violations Domestic violence Open container Disorderly conduct/public drunkenness Assault/battery/getnng into fights Criminal trespassing Minor in possession of alcohol Petty larceny Possession of illegal drugs OtherJ * Number of offenders divided by total number of study participants (n = 7). t Moving violations include reckless driving, running a red light, running a stop sign, speeding, signal required before turning, failure to keep proper lookout, failure to stay within traffic lane, failure to yield to emergency equipment, right turn from left lane, passing a school bus, failure to yield to approaching vehicle in intersection, and others. % Other offences include animal licence required, negligent use of firearm, concealing identity, criminal damage to property, malicious telephone calls, prostitution, vicious animals, rabies vaccinations, child abandonment, illegal fireworks, sterilization, illegal burning of refuse/waste, littering, assisting minor to obtain alcohol, leaving the scene of an accident, unreasonable noise/horn, permitting unauthorized person to drive, spotlight w/glaring beam, no child restraint in car seat, meter parking/expired meter, throwing glass/nails/tacks on highway, and others Validity of self-reported previous DWIs, criminal offences, and traffic violations Court records and LCSP records were compared for three attributes: previous DWIs, criminal offences and traffic violations (Tables -). Table 5 shows results for all three attributes combined. The results were tabulated into one of the following four categories: (1) no/no: no previous arrests either in court or LCSP records; () no/yes: no previous court records, but reported previous arrests to LCSP; () yes/no: previous court records, but reported no previous arrests to LCSP; () yes/yes: previous arrests in both court and LCSP records. Category () would give an estimate of underreporting due to lost records by Table b. Agreement* between court records and selfreports in the screening programme (based on those with court records) Screening programme n % Underre ported! Table a. Whether any previous DWIs reported by the court Same and the screening programme (based on all study Overreported participants) Court Screening programme * 5 6 * Comparisons were based on number of previous DWIs. f Underreported: offender's response was negative or partially negative to the screening programme but court records were affirmative. Overreported: offender's response was affirmative to the screening programme but court records were negative. X Thirty-eight participants with court records available (see Table a). Downloaded from on 19 August 018

5 SCREENING DWI OFFENDERS 587 Table a. Whether any previous criminal offences* reported by the court and the screening programme (based on all study participants) Table a. Whether any previous traffic violations reported by the court and the screening programme (based on all study participants) Court Screening programme * Previous criminal offences other than DWIs Court Screening programme Table b. Agreement* between court records and selfreports in the screening programme (based on those with court records) Screening programme n % Underreportedf Same Overreported Completely different * Comparisons were based on types of previous criminal offences. t Underreported: offender's response was negative or partially negative to the screening programme but court records were affirmative. Overreported: offender's response was affirmative to the screening programme but court records were negative. Completely different: offender's response contained offences not revealed in court records, while court records contained offences not revealed by offender. % 6 participants with court records available (see Table a) the court, plea bargaining with the judge, pre-198 court records, misreport of the clients, and arrests outside the jurisdiction. On the other hand, category () could yield a scale for estimated underreporting of previous offences in selfreports. For those with court records available (categories and ), the court and LCSP records were further compared to see if they were the same, overreported by the client, underreported by the client, or completely different. For previous DWIs (Table b), the comparison was made based on the number of DWI arrests. For criminal offences (Table b), the types of offences were used because no data on frequency were available. For traffic violations (Table b), the comparison was based on the number of arrests. For all three Table b. Agreement* between court records and selfreports in the screening programme (based on those with court records) Screening programme n % Underre portedj Same Overreported * Comparisons were based on number of previous traffic violations. f Underreported: offender's response was negative or partially negative to the screening programme but court records were affirmative. Overreported: offender's response was affirmative to the screening programme but court records were negative. % 10 participants with court records available (see Table a). attributes combined (Table 5b), 'underreported' means that any of the three offences were 'underreported'; 'same' means all three were the 'same'; and 'overreported' means any of the three were 'overreported'. In general, the category 'same' represents those who appeared to be honest. The category 'underreported' represents those who did not report or only gave partial information. The category 'overreported' represents additional records or arrests that occurred outside the jurisdiction, were lost by the court, plea-bargained with the judge or misreported by the clients. The category 'completely different' represents those who misreported completely or did not report any of the arrests that were in the court system. Most of the study participants (n = 0) were first-time DWI offenders; 80% of them had no previous DWI arrest records (Table a). Sixteen of these reported having been arrested for DWI, although these records were not retrievable from Downloaded from on 19 August 018

6 588 I. CHANG and S. C. LAPHAM Table 5a. Whether any previous offences reported by the court and the screening programme (based on all study participants) Court Screening programme n t % Table 5b. Agreement between court records and selfreports in the screening programme (based on those with court records) Screening programme n % Underre ported* Same Overreported Other Underreported: if any one of the attributes (DWI, traffic violation, criminal offence) was underreported to the screening programme. Overreported: if any one of the attributes (DWI, traffic violation, criminal offence) was oveiteported to the screening programme. Other: for those with inconsistent results and who could not be classified into the above three categories. t 1 participants with court records available (see Table 5a). during interview. Court records or self-reports indicated previous arrests for traffic violations for a majority (6%). Only 7% (n = 101, Table a) of the clients had no arrests for traffic violations in either the LCSP or the court records. Twenty-five per cent (n = 69) of respondents reported previous arrests for traffic violations, but showed no court records. Of those who had court records (n = 10, Table b), 58% (n = 61) underreported, indicating that they had not reported all or some of the previous offences during evaluation; 0% (n = 1) gave self-reports that agreed with the court records; and % (n = ) overreported. Twenty-five per cent (n = 69, Table 5a) of the clients indicated no arrests for any offence and had their statement supported by court records. Twenty-six per cent (n = 7), according to their LCSP reports, had other previous arrests, but those records were not available from the court system. Of those with court records (n = 1, 9%, Table 5b), a substantial majority (65%; n = 87) underreported, indicating that they had not reported all of their previous offences during evaluation; % {n- 18) had self-reports that agreed with court records; and 8% (n = 8) of them overreported. Fifty-seven per cent of the single offenders (5 out of 95) underreported, whereas 87% of the multiple offenders (5 out of 8) underreported. the court system. Of those who had court records (n = 8, Table b), almost half (%, n = 16) of them underreported, indicating that they had not reported all or some of the previous DWIs during evaluation, whereas 58% of them (n = ) said they had at least one DWI arrest, and court records corroborated this. Sixty-eight per cent of the offenders (n = 187) had no arrests for criminal offences in either the LCSP or the court (Table a) systems. Twentythree of them (8%) reported having been arrested for criminal offences other than DWI, although their records were not retrievable from the court system. Of those who had court records (n = 6, Table b), 66% (n = ) of them underreported, indicating that they had not reported all or some of the previous offences during evaluation; 1% (n = 8) of them were in agreement with the court; and % overreported. The 18% in the category 'completely different' also calls into question the accuracy of some data collected DISCUSSION The offence profile (Table 1) of this DWI population has been used by another study (Wells- Parker et al., 1986). The types of offences shown in these two studies are quite similar, even though the numbers of occurrences for various types of offences differed in these study populations. Differences were primarily due to differences in sample size, the multiple/single DWI offender ratio, and probably differences in local law enforcement activities. Researchers have long documented the deviant behaviours of the DWI population. Beerman et al. (1988) reported that prior criminal history, moving violations and BAC are the best predictors for recidivism. chajski et al. (199) also concluded that multiple DWI offenders are more likely to have repeated traffic tickets throughout their lives and a criminal history for crimes other than DWI. This study shows that three-quarters of the Downloaded from on 19 August 018

7 SCREENING DWI OFFENDERS 589 offenders have had at least one previous offence or traffic violation. Furthermore, multiple offenders tend to offend over and over. This study also estimates that 65% of the offenders underreported their arrests to their counsellors, presumably to avoid referral to treatment for alcohol-related problems. This incidence of underreporting is conservative, as it was derived from a population dominated by first-time offenders. For a population with a higher percentage of multiple offenders, such as -8% multiple DWI offenders in New Mexico (New Mexico State Highway and Transportation Department, 199), underreporting could well exceed 65%. An estimated 7% and 8% of the DWI screening facilities in rth Carolina, USA did not have access to driving records and court criminal records, respectively (Popkin, 1989). Therefore, direct access by the screening agencies to court records during DWI screening is strongly recommended. This study shows a higher percentage of underreporting than other studies (Sobell et ai, 197, Sobell and Sobell, 1975, 1978; Cooper et ai, 1980). Those studies, which also focused on criminal and driving records of the participants, were based on smaller populations of hospital in/ out-patients for treatment of alcoholism. Most of these studies showed moderate-to-good agreement between self-reported data and official records, widi the most common discrepancy being overreporting (Midanik, 198; Babor et ai, 1987). Several factors might have contributed to the differences between this study and others. First, participants in four of the five studies on alcoholic populations were volunteers and had no treatment consequences based on their self-reported data, whereas in this study, participants were mandated by the court to go through the evaluation for alcohol-related problems and faced consequences (court-ordered referral to treatment). This could account for the higher rate of underreporting in this study. Also, two out of the four other studies on alcoholic populations used a test-retest strategy to collect more reliable data, yielding greater consistency between self-reported data and official records. Other differences between this study and those mentioned above include the status of participants (DWI arrests vs alcoholic patients), the sample size (7 vs 1-70), and variations in analysis strategies. A significant portion of the offenders who did not have court records did report having been arrested. Such offenders appear to be honest. Various explanations could account for this discrepancy: (1) the arrests occurred outside the court's jurisdiction; () the court lost the records; () offenders plea-bargained with the judge; () offenders misreported; and (5) arrests occurred before 198, so were not reported by the court. Measures that could reduce these discrepancies would be to: record all dismissals of citations, as suggested by the National Transportation Safety Board (1987); build concurrently accessible drivers' history files, and increase interagency links between the court, the Police Department, the Motor Vehicle Department, and the Licensing Authority. Having access to court records is critical. So too is using a well-designed questionnaire (as discussed below) to obtain a more complete history on criminal offences and traffic violations through self-reporting during a screening process. From the standpoint of conducting a wellstructured interview, several researchers (Skinner, 198; Babor et ai, 1987) have suggested various procedures and techniques to minimize response bias and increase data validity. Examples of these techniques are: use of well-developed instruments, standardized interview protocols, memory aids, and bounded recall techniques. In addition, it is important for the counsellors and probation officers to know the predictors of the offence characteristics of drunk driving by the targeted DWI population, because this understanding could assist in guiding interviews and predicting future recidivism. Based on the offence profile established in this study, questions pertaining to the following are recommended to be part of the structured interview to obtain criminal or offence history: (1) alcohol-related arrests, such as previous DWIs, open container violations, minor in possession of alcohol, public drunkenness, and disorderly conduct; () criminal behaviours, such as domestic violence, assault or battery, obstructing or evading officers, and criminal trespassing; () hazardous driving behaviours, such as reckless driving, running a stop sign, running a red light, speeding, and other moving violations. These findings could help gauge the need for treatment for alcohol-related problems. In summary, strong evidence exists that DWI offenders undenreport their previous arrests during Downloaded from on 19 August 018

8 590 I. CHANG and S. C. LAPHAM DWI screening for alcohol-related problems. Because this information is critical for a client evaluation and recidivism prediction, court records should be made available to screening agencies. Court records are not always complete, as clearly demonstrated in the present study. Therefore, structured questions should also be included in client interviews to provide additional information about the client's offence profile, and assist counsellors in making alcohol abuse assessments. Acknowledgements The authors would like to thank Ms Donna Pohuma for her careful retrieval and review of court records, Ms Joyce Welt for word-processing support, and Mr Wilfrid Koponen for technical editing of the manuscript. This study was supported by the Lovelace Institutes and National Institute on Alcohol Abuse and Alcoholism, grant #R0l AA REFERENCES Babor, T. F., Stephens, R. S. and Marlatt, G. A. (1987) Verbal report methods in clinical research on alcoholism: Response bias and its minimization. Journal of Studies on Alcohol 8, 10-. Beerman, K. A., Smith, M. M. and Hall, R. L. (1988) Predictors of recidivism in DWIs. Journal of Studies on Alcohol 9, -9. Chang, I., Lapham, S. C. and Barton, K. J. (1996) Drinking environment and sociodemographic factors among DWI offenders. Journal of Studies on Alcohol 57, Cooper, A. M., Sobell, M. B., Maisto, S. A. and Sobell, L. C. (1980) Criterion intervals for pretreatment drinking measures in treatment evaluation. Journal of Studies on Alcohol 1, 86. Donovan, D. M., Queisser, H. R., Salzberg, P. M. and Umlaff, R. L. (1985) Intoxicated and bad drivers: subgroups within the same population of high-risk men drivers. Journal of Studies on Alcohol 6, Fuller, R. K., Lee, K. K. and Gordis, E. (1988) Validity of self-report in alcoholism research: results of a Veterans Administration cooperative study. Alcoholism: Clinical and Experimental Research 1, Hesselbrock, M., Babor, T. F., Hesselbrock V., Meyer, R. E., and Workman. K. (198) 'Never believe an alcoholic?' On the validity of self-report measures of alcohol dependence and related constructs. International Journal of Addictions 18, Lapham, S. C, Skipper, B. J., Owen, J. P., Kleyboecker, K. and Teaf, D. (1995) Alcohol abuse screening instruments: normative test data collected from a first DWI offender screening program. Journal of Studies of Alcohol 56, Mayhew, D. R., Beimess. D. J., Simpson, H. M. and Lamble, R. W. (199) Diagnostic assessment of problem drivers: state of assessment and treatment techniques. Ministry of Transportation, Ontario, June, 199. Midanik, L. (198) The validity of self-reported alcohol consumption and alcohol problems: a literature review. British Journal of Addiction 77, National Transportation Safety Board (1987) Deficiencies in enforcement, judicial, and treatment programs related to repeat offender drunk drivers. Alcohol, Drugs and Driving, 1-. New Mexico Criminal and Traffic Law Manual (199) The Michie Company, Charlottesville, Virginia. New Mexico State Highway and Transportation Department, Traffic Safety Bureau (199) Driving While Impaired in New Mexico. NIAAA (199) Alcohol and Health, The Seventh Special Report to the U.S. Congress on Alcohol and Health. National Institute on Alcohol Abuse and Alcoholism, Washington, DC. chajski, T. H., Miller, B.A. and Parks, K. A. (199) Comparison of first-time and repeat DWI offenders. Paper presented at the Annual Meeting of the Research Society on Alcoholism, Maui, Hawaii, 18- June 199. Polich J. M. (198) The validity of self-reports in alcoholism research. Addictive Behaviors 7, 1-1. Popkin, C. L. (1989) Mandatory Substance Abuse Assessments in rth Carolina (Final Report) (1989). University of rth Carolina Highway Safety Research Center, June Popkin, C. L., Kannenberg, C. H., Lacey, J. H., and Waller, P. F. (1988) Assessment of Classification Instruments Designed to Detect Alcohol Abuse (Final Report). US Dept of Transportation, National Highway Traffic Safety Administration, Washington, D.C., 0590 (DOT HS ), 1-. Skinner, H. A. (198) Assessing alcohol use by patients in treatment. In Research Advances in Alcohol and Drug Problems, Smart, R. J. A. and Cappell, H. D. eds. Plenum Press, New York. pp Sobell, L. C. and Sobell, M. B. (1975) Outpatient alcoholics give valid self-reports. Journal of Nervous and Mental Disease 161,. Sobell, L. C. and Sobell, M. B. (1978) Validity of selfreports in three populations of alcoholics. Journal of Consulting and Clinical Psychology 6, Sobell, M. B., Sobell, L. C. and Samuels, F. H. (197) Validity of self-reports of alcohol-related arrests by alcoholics. Quarterly Journal of Studies on Alcohol 5, Watson, C. G., Tilleskjor, C, Hoodecheck-Schow, E. A., Purcel, J. and Jacobs, L. (198) Do alcoholics give valid self-reports? Journal of Studies on Alcohol 5, -8. Wells-Parker, E., Cosby, P. J. and Landrum, J. W. (1986) A typology for drinking driving offenders: methods for classification and policy implications. Accident Analysis and Prevention 18, 5. Wilson, J. R. and Jonah, B. A. (1985) Identifying impaired drivers among the general driving population. Journal of Studies on Alcohol 6, Downloaded from on 19 August 018