This article was downloaded by: [UZH Hauptbibliothek / Zentralbibliothek Zürich] On: 25 December 2014, At: 15:12 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20
MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse Robert C. Colligan & Kenneth P. Offord Published online: 22 Jun 2011.
To cite this article: Robert C. Colligan & Kenneth P. Offord (1990) MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse, Journal of Personality Assessment, 55:3-4, 708-716, DOI: 10.1080/00223891.1990.9674106 To link to this article: http://dx.doi.org/10.1080/00223891.1990.9674106
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
JOURNAL OF PERSONALITY ASSESSMENT, 1990, 55(3&4), 708-716 Copyright @ 1990, Lawrence Erlbaum Associates, Inc.
MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse Robert C. Colligan Section of Psychology Mayo Clinic and Mayo Foundation
Kenneth P. Offord Section of Biostatistics Mayo Clinic and Mayo Foundation
MacAndrew (1965) constructed the MacAndrew Alcoholism Scale (MAC) using adult male alcoholic outpatients and male psychiatric outpatients. Later, he developed the Substance Abuse Proclivity Scale (SAP; MacAndrew, 1986),specifically for screening younger males, using adolescent and young adult substance misusing and nonmisusing males 16 to 22 years of age. Because clinicians frequently apply the MAC to adolescents, we compared the efficacy of the MAC and the SAP scales using a sample of 436 young males between the ages of 13 and 25, who were hospitalized because of substance abuse, and 770 normal young males between the ages of 13 and 25. Results suggest that clinicians should use these scales with great caution.
T h e enormous and adverse impact of chemical misuse o n adolescents, families, schools, and society in general is well known t o clinicians in medical and mental health settings. Some sense of the magnitude of the problem can be obtained from a review of the consecutive surveys of American high school students in various parts of the United States during the past 15 years by the University of Michigan Institute for Social Research. Data from this group indicate that, in a nationwide sample of approximately 7,700 high school senior girls and 7,100
709
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
SCREENING MALE ADOLESCENTS FOR SUBSTANCE MISUSE
boys, 2.8% of the girls and 6.7% ofthe boys reported alcohol use during 20 ofthe 30 days before the survey (Johnston, O'Malley, 6r Bachman, 1987). The incidence of alcohol and other substance misuse among adolescents exhibiting emotional or behavioral problems is even higher (Minnesota Department of' Education, 1989; Moore, 1984, 1988). The importance of screening adolescents and young adults for chemical misuse is obvious. The Minnesota Multiphasic Personality Inventory (MMPI) is' the most widely used and thoroughly researched of the objective ~ e r s o n a l i t ~ assessment questionnaires currently available (Lubin, Larsen, & Matarazzo, 1984; Lubin, Larsen, Matarazzo, & Seever, 1985; Piotrowski & Keller, 1989). The MacAndrew Alcoholism Scale (MAC; MacAndrew, 1965) is roba ably the: most widely used and thoroughly researched of the many alcoholism scales derived from the MMPI item pool (Preng & Clopton, 1986; Prescott 6. Gottesman, 1988). Although the MAC scale was developed by contrasting the MMM responses from 300 adult male alcoholic outpatients with 300 adult male nonalcoholic psychiatric outpatients (MacAndrew, 1965, 1981), it has also been applied to adolescents and young adults of both sexes with varying reports of usefulness. More specifically, some researchers have found the MAC scale to have potential value in substance-abuse screening for adolescents of both sexes (Wolfson & Erbaugh, 1984),even in a 20-item abbreviated form (Rathus, Fox, & Ortins, 1980) and to be a potential index of the severity of chemical misuse (Hinge, 1983; Moore, 1984), but it requires the application of cutting scores different from those used with adults (Davis, Colligan, Morse, & Offord, 1987; Wisniewski, Glenwick, 6r Graham, 1985; Wolfson & Erbaugh, 1984). Some indications for predictive utility have even been reported ( H o h a n n , Loper, 6. Kammeier, 19741, and suggested norms for use with adolescent mental health referrals have been provided (Watson, Harris, Johnson, & LaBeck, 1983). However, MacAndrew (1979) emphasized that the MAC scale was "only accidentally" an "early identification instrument" (p, B9), because it had not been constructed for that purpose. He implied that a new and more definitive scale for addiction proneness was under way. Subsequently, MacAndrew reported on the development of a new Substance Abuse Proclivity scale (SAP)for the MMPI. He (MacAndrew, 1986) described it as "a psychometric instrument for the early detection of problem-engendering alcohol and drug use in young men. . ." (p. 161). To develop the new SAP scale, MacAndrew used procedures comparable to those he had used for constructing the MAC scale. More specifically, MMPI responses were obtained from 191 adolescent and young adult men, ranging in age from 16 to 22 years, who were receiving inpatient or outpatient treatment for either alcohol or drug misuse. Item responses from this criterion group of "alcohol offenders" and "substance misusersn (p. 161) were contrasted with MMPI responses from three samples of males who were also 16 to 22 years of age
.
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
710
COLLIGAN AND OFFORD
(N = 4,011) described as (a) "putatively normal" (p. 161; i.e., college students), (b) nonpsychiatric medical outpatients (Swenson, Pearson, & Osborne, 1973), and (c) nonsubstance-misusing psychiatric outpatients. After a careful three-stage item analysis, 36 items composed the SAP scale; these are listed in Table 1. Of these items, 12 also fell on the MAC scale. However, to maintain the scale as an indirect measure, as he had done with the MAC, MacAndrew excluded six MMPI items described as "circular in content" from consideration because they were believed to refer directly to alcohol/ substance misuse or its possible consequences. The excluded items are: 215. 460. 466. 496.
I have used alcohol excessively. I have used alcohol moderately (or not at all). Except by a doctor's orders, I never take drugs or sleeping powders.
I have never seen things doubled (that is, an object never looks like two
objects to me without my being able to make it look like one object). 156. I have had periods in which I carried on activities without knowing later what I had been doing. 25 1. I have had blank spells in which my activities were interrupted and I did not know what was going on around me. When the final version of the SAP scale was applied to the criterion groups from the original derivation samples, detection rates ranged from 81% to 89%. Subsequently, MacAndrew obtained six additional samples for cross-validation. A raw score of 16 or more on the SAP scale identified 90% of the substance misusers and 86% of the nonmisusers among 412 subjects from the crossvalidation groups. The mean scores for the alcohol (M = 20.0; SD = 4.3) and the substance misusers (M = 20.1; SD = 4.3) from the original criterion samples were slightly lower than the mean scores obtained from the combined group of alcohol and other substance abusers in the cross-validation groups (M = 21.6; SD = 4.3), and the distribution of scores was described as virtually identical between the two sets of data. Also, the mean SAP scale scores among the nonchemical-misusing samples were consistent. Among the college students and medical and psychiatric outpatients in the contrast samples used for scale development, mean SAP scores ranged from 10.4 to 11.8 (SDs ranged from 3.7 to 4.0) and were very consistent with the mean score range (8.3 to 11.0) from the college students in the cross-validation samples.
PROBLEM The clinical applicability and utility of any new MMPI scale require that it be evaluated in various settings (MacAndrew, 1987, 1989). Therefore, both the SAP and the MAC scales were applied to a sample of 770
TABLE 1 Items of the MMPI SAP Scale Listed in Item Order by Direction of Endorsement for Scoring
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
True My sex life is sahsfactory. As a youngster I was suspended from school one or more times for cutting up. I am a good mixer. I have not lived the right kind of life. I believe in the second coming of Christ. I like to go to parties and other affairs where there is lots of loud fun. In school I was sometimes sent to the principal for cutting up. I know who is responsible for most of my troubles. When I was a chdd, I belonged to a crowd or gang that tried to stick together through thick and thin. I feel that I have often been punished without cause. I belleve I am a condemned person. I like to flirt. I think I would like the work of a building contractor. My parents have often objected to the kind of people I went around with. I don't blame anyone for trying to grab everything he can get in this world. If people had not had it in for me I would have been much more successful. I have certainly had more than my share of things to worry about. I have no enemies who really wish to harm me. I feel uneasy indoors. I played hooky from school quite often as a youngster. I like or have liked fishing very much. I have often found people jealous of my good ideas just because they had not thought of them first. In school my marks in deportment were quite regularly bad. I have one or more faults which are so big that ~tseems better to accept them and try to control them rather than to try to get rid of them. I am afraid of finding myself in a closet or small closed place. I have frequently worked under people who seem to have things arranged so that they get credit for good work but are able to pass off mistakes onto those under them. Several times a week I feel as if something dreadful is about to happen. False 137. I believe that my home life is as pleasant as that of most people I know. "173. I liked school. "179. I am worried about sex matters. "294. I have never been in trouble with the law. 376. Policemen are usually honest. 377. At parties I am more likely to sit by myself or with just one other person than to join in with the crowd. "378. I do not like to see women smoke. 464. I have never seen a vision. 532. I can stand as much pain as others can.
-
"An item thar is also included on the MMPI MAC scale and scored in the same direction (MacAndrew, 1965).
712
COLLIGAN AND OFFORD
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
normal male adolescents and young adults, ranging in age from 13 to 25 years (Colligan & Offord, 1989; Colligan, Osborne, Swenson, & Offord, 1989), and to a sample of 436 adolescent and young adult men, ages 13 to 25 years, who had been hospitalized for evaluation or treatment of substance misuse. Members of the normal contrast sample had been screened for potentially handicapping physical or mental conditions and for whether or not they were under a physician's care according to Hathaway's original procedures (Hathaway & McKinley, 1940; McKinley & Hathaway, 1940).
RESULTS The top portion of Table 2 provides the descriptive statistics for the SAP and the MAC scales among our samples of 770 normal young men and 436 substance misusers. It should be noted that the mean SAP scores for both groups are noticeably different from those reported by MacAndrew (1986). In our normal contrast sample, the mean raw score was 13.9 (SD = 4.3), whereas the range of mean values was 10 to 12 in the SAP derivation study. In addition, the mean score of 18.4 (SD = 4.1) in our substance-misusing group is noticeably lower than the mean raw score range of 20 to 22 in the SAP derivation study. Use of a raw score of 16 or more on the SAP, as recommended by MacAndrew (1986), correctly identified 74% of our criterion group of substance misusers, which was somewhat lower than the 81% to 90% reported by MacAndrew for the derivation and cross-validation samples he used. Furthermore, in addition to misclassifying 26% of our substance misusers, the SAP identified 35% of our normal contrast group as falling in the substance-misusing category. Stated conversely, in our samples, the SAP obtained a specificity of 65% and a sensitivity of 74%. The preceding information is insufficient, however, to appreciate how the SAP will perform in any given clinical setting because of varying degrees of prevalence of substance abuse/misuse. Therefore, it is important to assess efficacy in terms of predictive power (Baldessarini, Finklestein, & Arana, 1983; Glaros & Kline, 1988). Assuming a 74% sensitivity and 65% specificity for the SAP, it is useful to consider the positive predictive value (i-e., the percentage of cases correctly identified as substance misusers of all those who have been identified by the SAP as substance misusers). T o do the calculation, the prevalence of substance abuse in the population to be screened is needed. If we were to assume prevalence rates of 5%, lo%, 25%, and 50%, the corresponding positive predictive values are 11%, 19%, 4 1%, and 68%. If the prevalence in the population t o be screened is less than 5%, then the positive predictive value for the SAP would be even less than 1I%, a relatively dismal performance. Subsequently, the MAC scale was also applied to our two samples. We used the raw score of 26 or more as recommended by Wolfson and Erbaugh (1984)
TABLE 2 SAP and MAC Raw Score Distributions Among Male Adolescents and Young Adults Ages 13 to 25 Years Scale
-
MAC
SAP
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
n
M Median SD Range
Normals
Abusersa
Normals
770 13.9 13.5 4.3 4-28
436 18.4 18.5 4.1 6-3 1
770 20.3 20.0 3.8 10-33
Abusersa
436 23.1 23.0 3.9 13-34
% of Sample With Raw Scores < x
Raw Score (x) - -
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
0.0 0.4 1.0 2.3 4.8 8.7 14.7 24.2 33.0 41.4 50.0 57.7 64.9 72.2 78.2 84.9 89.9 93.5 95.6 96.6 98.3 98.8 99.4 99.7 99.9 100.0 100.0 100.0 100.0 100.0 100.0 100.0
aPersons who have abused drugs or alcohol.
0.0 0.0 0.0 0.2 0.2 0.7 1.6 2.8 5.0 7.6 11.7 19.5 25.7 31.0 39.0 50.0 59.6 69.5 77.3 84.6 89.2 94.3 96.8 97.9 98.9 99.5 99.8 99.8 100.0 100.0 100.0 100.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.9 1.9 3.4 6.0 9.6 16.1 23.8 32.6 40.0 51.2 62.3 71.9 82.2 86.2 92.2 94.8 96.5 98.7 99.2 99.7 99.9 99.9 100.0 100.0
0.Cl 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 1.1 2.3 4.1 6.9 11.21 18.11 27.5 36.;! 44.0 55.0 64.9 73.9 78.7 87.;! 91.5 94.7 97.7 98.4 98.0 99.8 100.0
-
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
714
COLLIGAN AND OFFORD
because their substance-misusing and contrast samples were obtained in the Midwest and descriptively seemed most comparable to our own. Examination of the descriptive statistics indicated that the means for our samples also differed noticeably from the values they reported. More specifically, in their normal contrast sample, the mean score on the MAC was 22 (SD = 4.2), whereas in our normal sample the mean value was 20 (SD = 3.8). The mean value was 27 (SD = 4.2) in their substance-misusing sample, whereas in ours it was 23 (SD = 3.9) on the MAC. When their cutting score was applied (26 or more), only 26% of our criterion group of substance misusers (sensitivity) were correctly identified, and when it was applied to our normal contrast sample, 92% were correctly identified (specificity) as nonsubstance misusers. MacAndrew (1965) originally'found a raw score of 24 or more on the MAC scale to be the most effective in classification. When this value was applied, the MAC identified 45% (sensitivity) of our substance-misusing subjects and correctly identified 82% of the normal contrast sample (specificity).
DISCUSSION In the samples just described, it is clear that the SAP was more effective than the MAC, consistent with MacAndrew's forecast. However, an evaluation of the positive predictive power for varying levels of prevalence still results in a significant risk for inaccurate classification. This is particularly troublesome because members of our substance-misusing group were all involved in inpatient treatment programs, and thus they seemed comparable to the adolescent and young adult samples used to derive the SAP. Our recent work in developing contemporary normal adolescent and adult samples for the MMPI (Colligan, Osborne, Swenson, & Offord, 1983, 1989; Colligan & Offord, 1987a, 198713, 1989) has demonstrated remarkable changes in MMPI response patterns since 1937, when work on the MMPI was initiated. The derivation and cross-validation samples used in the construction of the SAP were not concurrent. Rather, MMPI responses were gathered from samples obtained in "the late 1930s," "the early 1960s," data described in articles published during the 1970s, and five samples that were undated. Because clear changes in MMM response patterns have been documented among normal adolescents and adults, such changes are also possibly present among chemical misusers; these would thus account for some of the shrinkage in classification accuracy observed in our study. It could also be argued that some unknown proportion of our normal contrast subjects were actually misusing alcohol or other substances on the basis of the survey data reported by Johnston et al. (1987). Although participants in the normal contrast sample had been screened, unreported (or denied) substance misuse may have allowed subjects who were substance misusers to be partici-
SCREENING MALE ADOLESCENTS FOR SUBSTANCE MISUSE
715
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
pants. Thus, the specificity reported for the SAP and the, MAC may be an underestimate. However, the very modest degree of sensitivity for the SAP, despite its careful construction, and the disappointing sensitivity of the MAC, among members of our substance-misusing sample, suggest that extreme caution should be used when these scales are applied clinically, a caveat recenthi emphasized in the literature by Gottesman and Prescott (1989).
ACKNOWLEDGMENTS We express appreciation to Leo J. Davis, Jr., PhD, Consultant, Section of Psychology; Robert M. Morse, MD, Director, Addiction Disorders Service; and Lloyd A. Wells, PhD, MD, Head, Section of Child and Adolescent Psychiatry, for their review of an early draft of this article and their permission to use the adolescent and young adult patients who were chemical misusers, and to Jeffrey J. Larson-Keller for his assistance with the data analysis.
REFERENCES Baldessarini, R. J., Finklestein, S., & Arana, G. W. (1983). The predictive power of diagnostic tes,ts and the effect of prevalence of illness. Archives of General Psychiatry, 40, 569-573. Colligan, R. C., & Offord, K. P. (1987a). The MacAndrew Alcoholism Scale applied to a contemporary normative sample. lournal of Clinical Psychology, 43, 291-293. Colligan, R. C., & Offord, I
Journal of Personality Assessment Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hjpa20
MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse Robert C. Colligan & Kenneth P. Offord Published online: 22 Jun 2011.
To cite this article: Robert C. Colligan & Kenneth P. Offord (1990) MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse, Journal of Personality Assessment, 55:3-4, 708-716, DOI: 10.1080/00223891.1990.9674106 To link to this article: http://dx.doi.org/10.1080/00223891.1990.9674106
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
JOURNAL OF PERSONALITY ASSESSMENT, 1990, 55(3&4), 708-716 Copyright @ 1990, Lawrence Erlbaum Associates, Inc.
MacAndrew Versus MacAndrew: The Relative Efficacy of the MAC and the SAP Scales for the MMPI in Screening Male Adolescents for Substance Misuse Robert C. Colligan Section of Psychology Mayo Clinic and Mayo Foundation
Kenneth P. Offord Section of Biostatistics Mayo Clinic and Mayo Foundation
MacAndrew (1965) constructed the MacAndrew Alcoholism Scale (MAC) using adult male alcoholic outpatients and male psychiatric outpatients. Later, he developed the Substance Abuse Proclivity Scale (SAP; MacAndrew, 1986),specifically for screening younger males, using adolescent and young adult substance misusing and nonmisusing males 16 to 22 years of age. Because clinicians frequently apply the MAC to adolescents, we compared the efficacy of the MAC and the SAP scales using a sample of 436 young males between the ages of 13 and 25, who were hospitalized because of substance abuse, and 770 normal young males between the ages of 13 and 25. Results suggest that clinicians should use these scales with great caution.
T h e enormous and adverse impact of chemical misuse o n adolescents, families, schools, and society in general is well known t o clinicians in medical and mental health settings. Some sense of the magnitude of the problem can be obtained from a review of the consecutive surveys of American high school students in various parts of the United States during the past 15 years by the University of Michigan Institute for Social Research. Data from this group indicate that, in a nationwide sample of approximately 7,700 high school senior girls and 7,100
709
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
SCREENING MALE ADOLESCENTS FOR SUBSTANCE MISUSE
boys, 2.8% of the girls and 6.7% ofthe boys reported alcohol use during 20 ofthe 30 days before the survey (Johnston, O'Malley, 6r Bachman, 1987). The incidence of alcohol and other substance misuse among adolescents exhibiting emotional or behavioral problems is even higher (Minnesota Department of' Education, 1989; Moore, 1984, 1988). The importance of screening adolescents and young adults for chemical misuse is obvious. The Minnesota Multiphasic Personality Inventory (MMPI) is' the most widely used and thoroughly researched of the objective ~ e r s o n a l i t ~ assessment questionnaires currently available (Lubin, Larsen, & Matarazzo, 1984; Lubin, Larsen, Matarazzo, & Seever, 1985; Piotrowski & Keller, 1989). The MacAndrew Alcoholism Scale (MAC; MacAndrew, 1965) is roba ably the: most widely used and thoroughly researched of the many alcoholism scales derived from the MMPI item pool (Preng & Clopton, 1986; Prescott 6. Gottesman, 1988). Although the MAC scale was developed by contrasting the MMM responses from 300 adult male alcoholic outpatients with 300 adult male nonalcoholic psychiatric outpatients (MacAndrew, 1965, 1981), it has also been applied to adolescents and young adults of both sexes with varying reports of usefulness. More specifically, some researchers have found the MAC scale to have potential value in substance-abuse screening for adolescents of both sexes (Wolfson & Erbaugh, 1984),even in a 20-item abbreviated form (Rathus, Fox, & Ortins, 1980) and to be a potential index of the severity of chemical misuse (Hinge, 1983; Moore, 1984), but it requires the application of cutting scores different from those used with adults (Davis, Colligan, Morse, & Offord, 1987; Wisniewski, Glenwick, 6r Graham, 1985; Wolfson & Erbaugh, 1984). Some indications for predictive utility have even been reported ( H o h a n n , Loper, 6. Kammeier, 19741, and suggested norms for use with adolescent mental health referrals have been provided (Watson, Harris, Johnson, & LaBeck, 1983). However, MacAndrew (1979) emphasized that the MAC scale was "only accidentally" an "early identification instrument" (p, B9), because it had not been constructed for that purpose. He implied that a new and more definitive scale for addiction proneness was under way. Subsequently, MacAndrew reported on the development of a new Substance Abuse Proclivity scale (SAP)for the MMPI. He (MacAndrew, 1986) described it as "a psychometric instrument for the early detection of problem-engendering alcohol and drug use in young men. . ." (p. 161). To develop the new SAP scale, MacAndrew used procedures comparable to those he had used for constructing the MAC scale. More specifically, MMPI responses were obtained from 191 adolescent and young adult men, ranging in age from 16 to 22 years, who were receiving inpatient or outpatient treatment for either alcohol or drug misuse. Item responses from this criterion group of "alcohol offenders" and "substance misusersn (p. 161) were contrasted with MMPI responses from three samples of males who were also 16 to 22 years of age
.
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
710
COLLIGAN AND OFFORD
(N = 4,011) described as (a) "putatively normal" (p. 161; i.e., college students), (b) nonpsychiatric medical outpatients (Swenson, Pearson, & Osborne, 1973), and (c) nonsubstance-misusing psychiatric outpatients. After a careful three-stage item analysis, 36 items composed the SAP scale; these are listed in Table 1. Of these items, 12 also fell on the MAC scale. However, to maintain the scale as an indirect measure, as he had done with the MAC, MacAndrew excluded six MMPI items described as "circular in content" from consideration because they were believed to refer directly to alcohol/ substance misuse or its possible consequences. The excluded items are: 215. 460. 466. 496.
I have used alcohol excessively. I have used alcohol moderately (or not at all). Except by a doctor's orders, I never take drugs or sleeping powders.
I have never seen things doubled (that is, an object never looks like two
objects to me without my being able to make it look like one object). 156. I have had periods in which I carried on activities without knowing later what I had been doing. 25 1. I have had blank spells in which my activities were interrupted and I did not know what was going on around me. When the final version of the SAP scale was applied to the criterion groups from the original derivation samples, detection rates ranged from 81% to 89%. Subsequently, MacAndrew obtained six additional samples for cross-validation. A raw score of 16 or more on the SAP scale identified 90% of the substance misusers and 86% of the nonmisusers among 412 subjects from the crossvalidation groups. The mean scores for the alcohol (M = 20.0; SD = 4.3) and the substance misusers (M = 20.1; SD = 4.3) from the original criterion samples were slightly lower than the mean scores obtained from the combined group of alcohol and other substance abusers in the cross-validation groups (M = 21.6; SD = 4.3), and the distribution of scores was described as virtually identical between the two sets of data. Also, the mean SAP scale scores among the nonchemical-misusing samples were consistent. Among the college students and medical and psychiatric outpatients in the contrast samples used for scale development, mean SAP scores ranged from 10.4 to 11.8 (SDs ranged from 3.7 to 4.0) and were very consistent with the mean score range (8.3 to 11.0) from the college students in the cross-validation samples.
PROBLEM The clinical applicability and utility of any new MMPI scale require that it be evaluated in various settings (MacAndrew, 1987, 1989). Therefore, both the SAP and the MAC scales were applied to a sample of 770
TABLE 1 Items of the MMPI SAP Scale Listed in Item Order by Direction of Endorsement for Scoring
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
True My sex life is sahsfactory. As a youngster I was suspended from school one or more times for cutting up. I am a good mixer. I have not lived the right kind of life. I believe in the second coming of Christ. I like to go to parties and other affairs where there is lots of loud fun. In school I was sometimes sent to the principal for cutting up. I know who is responsible for most of my troubles. When I was a chdd, I belonged to a crowd or gang that tried to stick together through thick and thin. I feel that I have often been punished without cause. I belleve I am a condemned person. I like to flirt. I think I would like the work of a building contractor. My parents have often objected to the kind of people I went around with. I don't blame anyone for trying to grab everything he can get in this world. If people had not had it in for me I would have been much more successful. I have certainly had more than my share of things to worry about. I have no enemies who really wish to harm me. I feel uneasy indoors. I played hooky from school quite often as a youngster. I like or have liked fishing very much. I have often found people jealous of my good ideas just because they had not thought of them first. In school my marks in deportment were quite regularly bad. I have one or more faults which are so big that ~tseems better to accept them and try to control them rather than to try to get rid of them. I am afraid of finding myself in a closet or small closed place. I have frequently worked under people who seem to have things arranged so that they get credit for good work but are able to pass off mistakes onto those under them. Several times a week I feel as if something dreadful is about to happen. False 137. I believe that my home life is as pleasant as that of most people I know. "173. I liked school. "179. I am worried about sex matters. "294. I have never been in trouble with the law. 376. Policemen are usually honest. 377. At parties I am more likely to sit by myself or with just one other person than to join in with the crowd. "378. I do not like to see women smoke. 464. I have never seen a vision. 532. I can stand as much pain as others can.
-
"An item thar is also included on the MMPI MAC scale and scored in the same direction (MacAndrew, 1965).
712
COLLIGAN AND OFFORD
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
normal male adolescents and young adults, ranging in age from 13 to 25 years (Colligan & Offord, 1989; Colligan, Osborne, Swenson, & Offord, 1989), and to a sample of 436 adolescent and young adult men, ages 13 to 25 years, who had been hospitalized for evaluation or treatment of substance misuse. Members of the normal contrast sample had been screened for potentially handicapping physical or mental conditions and for whether or not they were under a physician's care according to Hathaway's original procedures (Hathaway & McKinley, 1940; McKinley & Hathaway, 1940).
RESULTS The top portion of Table 2 provides the descriptive statistics for the SAP and the MAC scales among our samples of 770 normal young men and 436 substance misusers. It should be noted that the mean SAP scores for both groups are noticeably different from those reported by MacAndrew (1986). In our normal contrast sample, the mean raw score was 13.9 (SD = 4.3), whereas the range of mean values was 10 to 12 in the SAP derivation study. In addition, the mean score of 18.4 (SD = 4.1) in our substance-misusing group is noticeably lower than the mean raw score range of 20 to 22 in the SAP derivation study. Use of a raw score of 16 or more on the SAP, as recommended by MacAndrew (1986), correctly identified 74% of our criterion group of substance misusers, which was somewhat lower than the 81% to 90% reported by MacAndrew for the derivation and cross-validation samples he used. Furthermore, in addition to misclassifying 26% of our substance misusers, the SAP identified 35% of our normal contrast group as falling in the substance-misusing category. Stated conversely, in our samples, the SAP obtained a specificity of 65% and a sensitivity of 74%. The preceding information is insufficient, however, to appreciate how the SAP will perform in any given clinical setting because of varying degrees of prevalence of substance abuse/misuse. Therefore, it is important to assess efficacy in terms of predictive power (Baldessarini, Finklestein, & Arana, 1983; Glaros & Kline, 1988). Assuming a 74% sensitivity and 65% specificity for the SAP, it is useful to consider the positive predictive value (i-e., the percentage of cases correctly identified as substance misusers of all those who have been identified by the SAP as substance misusers). T o do the calculation, the prevalence of substance abuse in the population to be screened is needed. If we were to assume prevalence rates of 5%, lo%, 25%, and 50%, the corresponding positive predictive values are 11%, 19%, 4 1%, and 68%. If the prevalence in the population t o be screened is less than 5%, then the positive predictive value for the SAP would be even less than 1I%, a relatively dismal performance. Subsequently, the MAC scale was also applied to our two samples. We used the raw score of 26 or more as recommended by Wolfson and Erbaugh (1984)
TABLE 2 SAP and MAC Raw Score Distributions Among Male Adolescents and Young Adults Ages 13 to 25 Years Scale
-
MAC
SAP
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
n
M Median SD Range
Normals
Abusersa
Normals
770 13.9 13.5 4.3 4-28
436 18.4 18.5 4.1 6-3 1
770 20.3 20.0 3.8 10-33
Abusersa
436 23.1 23.0 3.9 13-34
% of Sample With Raw Scores < x
Raw Score (x) - -
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
0.0 0.4 1.0 2.3 4.8 8.7 14.7 24.2 33.0 41.4 50.0 57.7 64.9 72.2 78.2 84.9 89.9 93.5 95.6 96.6 98.3 98.8 99.4 99.7 99.9 100.0 100.0 100.0 100.0 100.0 100.0 100.0
aPersons who have abused drugs or alcohol.
0.0 0.0 0.0 0.2 0.2 0.7 1.6 2.8 5.0 7.6 11.7 19.5 25.7 31.0 39.0 50.0 59.6 69.5 77.3 84.6 89.2 94.3 96.8 97.9 98.9 99.5 99.8 99.8 100.0 100.0 100.0 100.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.9 1.9 3.4 6.0 9.6 16.1 23.8 32.6 40.0 51.2 62.3 71.9 82.2 86.2 92.2 94.8 96.5 98.7 99.2 99.7 99.9 99.9 100.0 100.0
0.Cl 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 1.1 2.3 4.1 6.9 11.21 18.11 27.5 36.;! 44.0 55.0 64.9 73.9 78.7 87.;! 91.5 94.7 97.7 98.4 98.0 99.8 100.0
-
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
714
COLLIGAN AND OFFORD
because their substance-misusing and contrast samples were obtained in the Midwest and descriptively seemed most comparable to our own. Examination of the descriptive statistics indicated that the means for our samples also differed noticeably from the values they reported. More specifically, in their normal contrast sample, the mean score on the MAC was 22 (SD = 4.2), whereas in our normal sample the mean value was 20 (SD = 3.8). The mean value was 27 (SD = 4.2) in their substance-misusing sample, whereas in ours it was 23 (SD = 3.9) on the MAC. When their cutting score was applied (26 or more), only 26% of our criterion group of substance misusers (sensitivity) were correctly identified, and when it was applied to our normal contrast sample, 92% were correctly identified (specificity) as nonsubstance misusers. MacAndrew (1965) originally'found a raw score of 24 or more on the MAC scale to be the most effective in classification. When this value was applied, the MAC identified 45% (sensitivity) of our substance-misusing subjects and correctly identified 82% of the normal contrast sample (specificity).
DISCUSSION In the samples just described, it is clear that the SAP was more effective than the MAC, consistent with MacAndrew's forecast. However, an evaluation of the positive predictive power for varying levels of prevalence still results in a significant risk for inaccurate classification. This is particularly troublesome because members of our substance-misusing group were all involved in inpatient treatment programs, and thus they seemed comparable to the adolescent and young adult samples used to derive the SAP. Our recent work in developing contemporary normal adolescent and adult samples for the MMPI (Colligan, Osborne, Swenson, & Offord, 1983, 1989; Colligan & Offord, 1987a, 198713, 1989) has demonstrated remarkable changes in MMPI response patterns since 1937, when work on the MMPI was initiated. The derivation and cross-validation samples used in the construction of the SAP were not concurrent. Rather, MMPI responses were gathered from samples obtained in "the late 1930s," "the early 1960s," data described in articles published during the 1970s, and five samples that were undated. Because clear changes in MMM response patterns have been documented among normal adolescents and adults, such changes are also possibly present among chemical misusers; these would thus account for some of the shrinkage in classification accuracy observed in our study. It could also be argued that some unknown proportion of our normal contrast subjects were actually misusing alcohol or other substances on the basis of the survey data reported by Johnston et al. (1987). Although participants in the normal contrast sample had been screened, unreported (or denied) substance misuse may have allowed subjects who were substance misusers to be partici-
SCREENING MALE ADOLESCENTS FOR SUBSTANCE MISUSE
715
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 15:12 25 December 2014
pants. Thus, the specificity reported for the SAP and the, MAC may be an underestimate. However, the very modest degree of sensitivity for the SAP, despite its careful construction, and the disappointing sensitivity of the MAC, among members of our substance-misusing sample, suggest that extreme caution should be used when these scales are applied clinically, a caveat recenthi emphasized in the literature by Gottesman and Prescott (1989).
ACKNOWLEDGMENTS We express appreciation to Leo J. Davis, Jr., PhD, Consultant, Section of Psychology; Robert M. Morse, MD, Director, Addiction Disorders Service; and Lloyd A. Wells, PhD, MD, Head, Section of Child and Adolescent Psychiatry, for their review of an early draft of this article and their permission to use the adolescent and young adult patients who were chemical misusers, and to Jeffrey J. Larson-Keller for his assistance with the data analysis.
REFERENCES Baldessarini, R. J., Finklestein, S., & Arana, G. W. (1983). The predictive power of diagnostic tes,ts and the effect of prevalence of illness. Archives of General Psychiatry, 40, 569-573. Colligan, R. C., & Offord, K. P. (1987a). The MacAndrew Alcoholism Scale applied to a contemporary normative sample. lournal of Clinical Psychology, 43, 291-293. Colligan, R. C., & Offord, I
