Inventory of outcome expectancy in marijuana users (Inventário de Expectativas de Resultados em Usuários de Maconha, IERUM): development and validation

Abstracts

Introdução: Este estudo apresenta a construção e validação do Inventário de Expectativas de Resultados em Usuários de Maconha (IERUM), o qual avalia as expectativas de resultados em relação ao uso de maconha, podendo ser importante no tratamento e prognóstico dos dependentes químicos dessa substância. Método: O estudo foi realizado em uma amostra de 181 sujeitos usuários de maconha, homens (n = 145) e mulheres (n = 36), entre 18 e 55 anos (28,34±8,57); 118 estavam internados em unidade de dependência química, 23 em tratamento ambulatorial, 25 em grupo de autoajuda (narcóticos anônimos, alcoólatras anônimos) e 15 não estavam em tratamento. A escala de 17 itens foi criada a partir da experiência em dependência química e em validação de escalas, sendo realizada primeiramente a validação semântica. Resultados: De acordo com a análise fatorial, o IERUM dividiu-se em cinco fatores (aspectos emocionais, percepção, craving, aspectos cognitivos e sexualidade) que responderam por 60,18% da variância. Conclusão: O IERUM apresentou bons resultados psicométricos, podendo avaliar as expectativas de resultados relacionadas ao uso de maconha.

Cannabis; expectativas; IERUM; construção; validação

ORIGINAL ARTICLE

Inventory of Outcome Expectancy in Marijuana Users (Inventário de Expectativas de Resultados em Usuários de Maconha, IERUM): development and validation

Rosemeri Siqueira Pedroso^I; Maria da Graça Tanori de Castro^II; Renata Brasil Araujo^III

^IMestre, Psicologia Clínica, Pontifícia Universidade Católica do Rio Grande do Sul (PUCRS), Porto Alegre, RS, Brazil. Psicóloga, Doutoranda, Ciências Médicas, Psiquiatria, Hospital de Clínicas de Porto Alegre (HCPA), Centro de Pesquisa em Álcool e Drogas (CPAD), Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, RS, Brazil.

^IIMédica psiquiatra. Mestre, Psicologia Clínica, PUCRS. Professora, Curso de Pós-Graduação em Psiquiatria, Instituto Abuchaim, Porto Alegre, RS, Brazil.

^IIIPsicóloga. Doutora, Psicologia, PUCRS. Coordenadora, Programa de Dependência Química, Hospital Psiquiátrico São Pedro, Porto Alegre, RS, Brazil.

Correspondence

ABSTRACT

Introduction: This study presents the development and validation of the Inventory of Outcome Expectancy in Marijuana Users (Inventário de Expectativas de Resultados em Usuários de Maconha, IERUM), which evaluates the outcome expectancy regarding marijuana use, and may be useful in the treatment and prognosis of marijuana-dependent individuals.

Method: The sample comprised 181 male (n = 145) and female (n = 36) marijuana users. Their age ranged from 18 to 55 years old (28.34±8.57). One hundred and eighteen patients were hospitalized at a chemical dependency unit, 23 were being treated in an outpatient clinic, 25 were receiving treatment in a self-help group (drug addicts anonymous, alcoholic anonymous), and 15 were not being treated. The 17-item scale was developed based on the experience related to chemical dependency and scale validation, with the semantic validation being the first phase of the process.

Results: According to the factor analysis, the IERUM was divided into five factors (emotional aspects, perception, craving, cognitive aspects, and sexual aspects) that accounted for 60.18% of the variance.

Conclusion: The IERUM showed good psychometric properties and can be used to evaluate the outcome expectancy related to marijuana use.

Keywords:Cannabis, expectancy, IERUM, development, validation.

INTRODUCTION

Outcome expectancies regarding the effects of a psychoactive substance may trigger addictive behavior, leading individuals to decide to consume the substance in anticipation of the pleasure associated with its use. Cognition, culture and personality all influence one's beliefs in the effects of a certain drug, making one more vulnerable to initiation of psychoactive substance use.^1,2 Information about psychoactive substances obtained from the media, friends, and parent figures can also lead one to observe and learn certain behaviors, in the belief that the use of these substances is associated with positive or negative outcomes.^3,4

In the treatment of substance dependence, positive outcome expectancy must be analyzed very carefully as a factor in the relapse process, as a patient's beliefs regarding the effects of a certain psychoactive substance may lead him or her to actively seek out use of the substance, regardless of the actual effects of the drug and the extent to which they differ from its expected effects. As early as the 1980s, some scholars considered the "outcome expectancy" construct as a cognitive mediator of behavior, and particularly of substance use and dependence.⁵ The relevance of this subject, it is imperative that an instrument be developed and validated to serve as a specific measure of outcome expectancy toward drug use, including marijuana use.^2,6

Hawkins⁷ tested a series of models mediating the relation between personality-related affective variables and marijuana use in a sample of college students who had used marijuana on at least once instance. Results showed the high levels of emotional and psychological anxiety and expectancy of relaxation and stress relief were significantly associated with marijuana use. Outcome expectancies associating marijuana use with expanded consciousness, heightened perception, and improved cognition were also found. The highest levels of perceived expectancy were related to social and sexual facilitation effects. The study also suggested that emotional and psychological aspects are predictive of conformity motives for marijuana use.

The implications of marijuana use were assessed in a sample of 82 in adolescents who had high rates of alcohol, marijuana, and nicotine use and were undergoing treatment. Marijuana and nicotine use were associated with high rates of depressive symptoms. Adolescents expected marijuana to have a greater positive impact and very little negative impact on their lives, and were not confident in their ability to resist marijuana use, although many appeared to be interested in changing their substance use.⁸

One instrument for assessment of outcome expectancies in marijuana use, the Marijuana Expectancy Questionnaire (MEQ), has already been validated in Brazil.³ Validation was carried out in a sample of 400 participants (217 male, 183 female) aged 12 to 60, which included marijuana users (n = 147) and non-users (n = 253). Of the 147 marijuana users, 90 were considered chemically dependent, and 56 of these were receiving inpatient treatment. Cronbach's alpha was 0.89 for the full 78-item scale and 0.89 for the reduced, 55-item instrument. In the Brazilian sample, the 78-item scale was administered, but statistical analysis was performed on the reduced scale, as the psychometric properties of both were similar. The discriminant validity of the reduced scale was found to be statistically significant. Temporal stability was assessed in a sample of 123 participants with a one-week test-retest interval, and showed consistency in response patterns over time. As many items in the scale had to be excluded - and it nonetheless remained long - and factor analysis was rendered impossible by the psychometric limitations of the scale itself, a new scale was devised for assessment of the outcome expectancy construct as it applies to marijuana use: the Inventory of Outcome Expectancy in Marijuana Users (Inventário de Expectativas de Resultados em Usuários de Maconha, IERUM).

Construction of a measurement instrument requires the establishment of definitions to contextualize the target construct within the existing theoretical framework and the available literature. An instrument devised to measure a construct will not always manage to encompass the entire semantic range of a concept. In this study of the IERUM, we prioritized clear conceptual definitions that allow assessment of inventory quality, as the development and validation of a research instrument must be undertaken judiciously, and conceptual equivalence must be assured by means of brainstorming and review by expert raters.^9-11

The present article thus seeks to report the development and semantic and psychometric validation of the IERUM instrument.

MÉTODO

Study design and participants

This was a cross-sectional study using a convenience sample of 181 subjects. Of these, 150 (82.87%) were marijuana-dependent and 31 (17.13%), engaged in harmful use of marijuana, according to ICD-10 definitions.¹² The sample included both male (n = 145) and female (n = 36) participants, aged 18 to 55 (mean, 28.34±8.57); 118 were receiving inpatient treatment at chemical dependency units, 23 were receiving treatment in outpatient clinics, 25 were members of twelve-step groups (Narcotics Anonymous, Alcoholics Anonymous), and 15 were receiving no treatment. Formal education through the fifth-grade level was established as a criterion for inclusion. Subjects were excluded from the study when they did not meet this requirement, as were subjects who presented with any cognitive involvement that would jeopardize understanding or completion of the instrument.

All ethical considerations pertaining to human subject research were followed, and the project was approved by the Research Ethics Committee of the Hospital Psiquiátrico São Pedro, Porto Alegre, Rio Grande do Sul.

Instruments

The following research instruments were used in our study:

- Structured interviews seeking to establish a social and demographic profile for the study sample, a diagnosis of psychoactive substance use according to International Statistical Classification of Diseases and Related Health Problems (ICD-10) criteria,¹² and obtain a history of marijuana and other drug use.

- IERUM: a structured Likert scale (a type of psychometric scale commonly used in questionnaires¹³) composed of 17 items related to marijuana use. Items are distributed into five subscales (emotional aspects, perception, sexual aspects, cognitive aspects, and craving) and each is assigned a score from 1 (completely disagree) to 7 (completely disagree).

- Visual analog scales (VAS): widely used in the research setting.^14-16 This study used five scales, meant to assess emotional aspects, perception, sexual aspects, cognitive aspects, and craving.

Procedures

Development

Development of the IERUM was based on the following construct-defining categories: emotional aspects, perception, craving, cognitive aspects, and sexual aspects. This choice of categories was based on the range of criteria assessed in studies of outcome expectancy in marijuana use.^6-8 Development of the scale followed a series of steps: (a) analysis of the content dimensions of each subscale to permit definition of a content matrix; (b) development of 18 scale items, based on these definitions, on a review of the relevant literature on the outcome expectancy construct, and on analysis of another instrument previously validated for the Brazilian reality⁶; (c) a discussion on item assessment and selection based on wording clarity and precision, simplicity, and adequacy of the classification to the defined concepts; (d) randomization of items into descriptive statements reporting experiences related to marijuana use outcome expectancy, on a seven-point Likert-type scale anchored at "completely disagree" and "completely agree"; (e) revision of the wording of each item by a language expert; (f) a pilot study with 30 marijuana users, meant to assess understanding of the scale's instructions and statements; (g) brainstorming: an exploratory method that involves pooling ideas from a group of people in an attempt to obtain the best solutions for a certain issue.¹⁷ In this case, three psychologists specializing in the management of chemical dependency discusses the clarity and relevance of all items and helped revise, draft, and proofread the scale. The final version of the scale comprised 17 items. The scale development criteria set forth by Pasquali were followed throughout the process.⁹

Data collection

The final version of the scale was administered individually to each participant. After being made aware of the objectives of the study and the use of information obtained from the instrument and being assured that confidentiality would be maintained, participants signed a Free and Informed Consent form and filled out the study instrument themselves. Data collection occurred between January 2007 and July 2008.

Data analysis

All information collected was entered into the Statistical Package for the Social Sciences version 16.0 software package (SPSS Inc., Chicago, EUA). Exploratory analysis consisted of descriptive statistics, including frequencies. In validating the MEQ, Cronbach's alpha was used to measure internal consistency of the instrument, and Pearson's correlation coefficient to assess its stability over time. The Kolmogorov-Smirnov test was used to test for normality. The Mann-Whitney U test, factor analysis with promax rotation, and Pearson's correlation coefficient were employed in inferential analysis. The appropriateness of using factor analysis on IERUM data was assessed with the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett's test. The significance level was set at 0.05.

RESULTS

Socio-demographic profile

Of the 181 subjects in the sample, 150 were marijuana-dependent (82.87%) and 31 engaged in harmful use of marijuana (17.13%). On average, participants smoked 23.41 cannabis cigarettes ("joints") per week (DS, 32.62; range, 1-210].

The remaining socio-demographic data and information on psychoactive substance use patterns are shown in Table 1.

Semantic validation

Construction of the IERUM followed Pasquali's decision criteria for item relevance.⁹ Semantic validation of the instrument followed the steps described by Pedroso et al.¹⁰

Final version of the instrument

The IERUM initially comprised 18 items; however, one item (question 4) was excluded after factor analysis. The remaining items were distributed into five subscales: I) emotional aspects (questions 7, 14, 15, 16, and 18); II) perception (questions 2, 3, 5, 6, and 12); III) sexual aspects (questions 9 and 17); IV) cognitive aspects (questions 8, 11, and 13); and V) craving (questions 1 and 10).

Internal consistency

Estimated reliability for the whole 17-item scale, based on Cronbach's alpha, was 0.779.

Cronbach's alpha values for the five subscales were: emotional aspects, 0.42 (n = 5); perception, 0.67 (n = 5); sexual aspects, 0.83 (n = 2); cognitive aspects, 0.69 (n = 3); and craving, 0.69 (n = 2).

Factor analysis

The Kaiser-Meyer-Olkin measure of sampling adequacy displayed a value of 0.74 and Bartlett's test was significant (p < 0.001), showing that analysis of variance (ANOVA) would be an adequate method for validation of the scale.

Table 2 shows the distribution of items into scale factors and categories after promax rotation. The number of factors was not statistically predetermined; five factors was the result found, not merely confirmed, on factor analysis.

We used the criteria set forth by Tiffany & Drobes¹⁸ for assignment of items to factors: items were considered to belong to a factor if loaded ? 0.40 on one factor and < 0.25, on another, and if the difference in measure of sampling adequacy loading for the two factors was at least 0.20.

Factor I (emotional aspects) had an eigenvalue of 4.06 (explained variance, 22.58%); factor II (perception) had an eigenvalue of 3.04 (explained variance, 16.91%); factor III (sexual aspects) presented an eigenvalue of 1.41 (7.84% variance); f actor IV (cognitive aspects) had an eigenvalue of 1.22 (explained variance, 6.78%); and factor V (craving) had an eigenvalue of 1.09 for 6.07% variance. Total variance explained was 60.19%.

Question 4 was excluded from the instrument, as it did not meet the Tiffany & Drobes criteria¹⁸; the final version of the IERUM was therefore left with 17 items.

Criterion validity

Criterion validity was measured by assessing the correlation among subscales with Pearson's coefficient. All values were < 0.80, indicating no item redundancy, as set forth by Ferrans & Powers.¹⁹. Our findings are shown in Table 3.

Convergent validity

Convergent validity was assessed by computing Pearson's coefficient of correlation between the IERUM subscales and the visual analog scales. Results are shown in Table 4.

Discriminant validity

Discriminant validity was measured by means of the Mann-Whitney U, comparing IERUM subscales between marijuana-dependent participants and those who were not considered dependent but engaged in harmful use. Results are shown in Table 5.

Marijuana use outcome expectancies

Mean scores for each subscale were: emotional aspects, 24.24 (SD 6.39, range 5-35); perception, 20.13 (SD 8.11, range 5-35); sexual aspects, 7.60 (SD 4.70, range 2-14); cognitive aspects, 14.75 (SD 5.72, range 3-21); and craving, 5.53 (SD 4.16, range 2-14). Dividing these scores by the number of items in each subscale gives a mean item score of 4.85 for the emotional aspects scale, 4.03 for the perception scale, 3.80 for the sexual aspects scale, 4.92 for the cognitive aspects scale, and 2.77 for the craving scale.

DISCUSSION

After semantic validation, the IERUM proved to be an adequate and quite reliable instrument that can be administered easily and quickly. It also showed excellent semantic and conceptual equivalence, as shown during brainstorming.^10,11 The entire process was based on the work of Pasquali.⁹

Cronbach's alpha for the entire 17-item scale (that is, the final, post-factor analysis version of the IERUM) was 0.779, which is a perfectly satisfactory level of internal consistency according to Rowland et al.,²⁰ who recommended a value of 0.70 as the cutoff from which items should be considered to assess the overall construct in a consistent manner. Comparison of subscales in terms of internal consistency showed divergent results, which were more satisfactory for the craving (0.69, n = 2), cognitive aspects (0.69, n = 3), and sexual aspects (0.83, n = 2) subscales.

Factor analysis led to the exclusion of one item (question 4), which appeared to belong to two factors. The remaining 17 items were distributed across five subscales: I, emotional aspects; II, perception; III, cognitive aspects; IV, craving; and V, sexual aspects.

In terms of criterion validity, most correlations were positive and significant, with no evidence of item redundancy, as recommended by Ferrans & Powers.¹⁷ Absence of correlation was found only between the craving and sexual aspects, emotional and cognitive aspects, and cognitive and sexual aspects subscales.

On convergent analysis, we found positive but weak-to-moderate correlations, with particular emphasis on the correlations between the cognitive aspects subscale of the IERUM and the perception/cognitive aspects VAS and between the craving subscale of the IERUM and the craving VAS. The weak correlation between the perception subscale of the IERUM and the perception/cognitive aspects VAS may have been due a certain amount of overlap-the latter VAS encompasses certain features of the cognitive aspect subscale of the IERUM. We also found a positive but weak correlation between the sexual aspects scale of the IERUM and the sexuality VAS. These findings may be due to the low specificity and sensitivity of visual analog scales.¹⁵

On discriminant analysis, however, administration of the IERUM showed no significant difference between the beliefs of marijuana-dependent participants and those of harmful users. One may infer that this is because both groups used the substance, thus homogenizing their outcome expectancies. This finding differs from the results of discriminant analysis for validation of the MEQ,⁵ during which differences in questionnaire scores were found between marijuana users and non-users.

The highest mean scores per question were found, in decreasing order, in the emotional aspects, cognitive aspects, and perception subscales. This is in line with the results of some prior studies,^5-7 which stress the role of these aspects in marijuana use outcome expectancy. Lower mean scores for the craving scale were also found by Pedroso et al in their validation of the MEQ.⁶

An understanding of outcome expectancies can help therapists motivate marijuana-dependent individuals to quit^1-6 and thus serve as a guiding principle in the psychoeducational process of modifying misguided beliefs about the effects of the substance. Therapists and researchers must remember that cravings are triggered by positive outcome expectancies, which may lead to relapse.^2,16

The development of a brief scale is important in light of the cost reduction benefit this entails; shorter administration time also makes it easier to use the scale in the research setting and during psychotherapeutic interactions.

Among the limitations of this study, we must stress that the instrument was administered to only one group of subjects, all of whom were marijuana users (marijuana-dependent or engaging in harmful use of the substance); this interfered with the results of the discriminant analysis portion of the study. Subject profile was also a noteworthy limitation; most participants were inpatients, dependent on other psychoactive substances, and smoked a considerably high number of marijuana cigarettes per week.

We expect IERUM will prove useful not only in the research setting, but also in helping professionals who treat chemically dependent persons assess outcome expectancy in marijuana users, thus making the difficult management of dependency-and maintenance of abstinence-a bit less complicated.

REFERENCES

• 1. Eiser JR. Attitudes and beliefs. In: Baum A, McManus C, Newman S, Weinman J, West R, eds. Cambridge handbook of psychology, health and medicine. Cambridge: Cambridge University Press; 1997. p. 3-7.
• 2. Marlatt A, Gordon J. Prevenção da recaída: estratégia e manutenção no tratamento de comportamentos aditivos. Porto Alegre: Artes Médicas; 1993.
• 3. Pedroso RS, Oliveira MS, Araujo RB, Castro MG, Melo WV. Expectativas de resultados frente ao uso de álcool, maconha e tabaco. Rev Psiquiatr RS. 2006;28(2):198-206.
• 4. Brown SA. Drug effect expectancies and addictive behavior change. Exp Clin Psychopharmacol. 1993;1(1-4):55-67.
• 5. Stacy AW, Widaman KE, Marllat GA. Expectancy models of alcohol use. J Pers Soc Psychol. 1990;58(5):918-29.
• 6. Pedroso RS, Oliveira MS, Moraes JF. Tradução, adaptação e validação da versão brasileira da escala Marijuana Expectancy Questionnaire. Cad Saude Publica. 2007;23(1):63-73.
• 7. Hawkins LW. The impact of personality and effect on college student's motives for marijuana use. Dissertation Abstracts International: Section B: The-Sciences-and-Engineering. 2007;68:1306.
• 8. Stein LA, Colby S, Barnett N, Monti P, Lebeau-lebeau-Craven R, Golembeske C. Assessment of substance use and treatment implications for incarcerated teens. In: Eisenberg RM, Geller EB, eds. Proceedings from the 65th annual meeting of the college on problems of drug dependence. Bal Harbour; 2003.
• 9. Pasquali L. Princípios de elaboração de escalas psicológicas. Rev Psiquiatr Clin. 1998;25(5):206-13.
• 10. Pedroso RS, Oliveira MS, Araujo RB, Moraes JF. Tradução, equivalência semântica e adaptação cultural do Marijuana Expectancy Questionnaire (MEQ). Psico USF. 2004;9(2):129-36.
• 11. Duarte PS, Miyazaki MC, Ciconelli RM, Sesso R. Tradução e adaptação cultural do instrumento de avaliação de qualidade de vida para pacientes renais crônicos (KDQOL-SF TM). Rev Assoc Med Bras. 2003;49(4):375-81.
• ¹²
  Organização Mundial de Saúde. Classificação de transtornos mentais e de comportamento da CID-10: descrições clínicas e diretrizes diagnósticas. Porto Alegre: Artmed; 1993.
• 13. Likert R. A technique for the measurement of attitudes. Arch Psychol. 1932;140(22):1-55.
• 14. Araujo RB, Oliveira MS, Moraes JF, Pedroso RS, Port F, Castro MG. Validação da versão brasileira do Questionnaire of Smoking Urges-Brief. Rev Psiquiatr Clin. 2007;34(4):166-75.
• 15. Araujo RB, Oliveira MS, Mansur MA. A validação brasileira do Questionnaire of Smoking Urges. Cad. Saude Publica. 2006;22(10):2157-67.
• 16. Araujo RB, Oliveira MS, Pedroso RS, Miguel AC, Castro MGT. Craving e dependência química: conceito, avaliação e tratamento. J Bras Psiquiatr. 2008;57(1):57-63.
• 17. Osborn AF. Applied imagination: principles and procedures of creative thinking. New York: Charles Scribners Sons; 1953.
• 18. Tiffany ST, Drobes DJ. The development and initial validation of a questionnaire on smoking urges. Br J Addict. 1991;86(11):1467-76.
• 19. Ferrans CE, Powers MJ. Psychiatric assessment of the Quality of Life Index. Res Nurs Health. 1992;15:29-38.
• 20. Rowland D, Arkkelin D, Crisler L. Computer-based data analysis:using SPSSXI in the Social and Behavioral Sciences. Chicago: Nelson-Hall; 1991.

Publication Dates

• Publication in this collection
  05 Aug 2010
• Date of issue
  2010

History

• Accepted
  02 Oct 2009
• Received
  28 June 2009