The involvement of adolescents in offending activities is currently a prominent social problem because it has negative emotional, physical and economic effects throughout society (Bochenek and Delgado Reference Bochenek and Delgado2006; United Nations Department of Economic and Social Affairs 2004). Besides that, it is a very common phenomenon (Le Blanc Reference Le Blanc2003). Currently, in Brazil, there are about 90,000 adolescents judicialized. Therefore, it is important to know who these adolescents are in terms of psychological and social aspects, and in which way the justice system has to treat them – considering their psychosocial needs. As this phenomenon is very complex, it is necessary to choose a method to investigate this question and propose ways of dealing with it. As is known, there are three levels of analysis in criminology: the criminality/delinquency, the crime/offense and the criminal/the delinquent. Each one of these levels, according to Pinatel (Reference Pinatel1963) and Rico (Reference Rico1977), has its own perspectives, ways of reasoning and methods. In the scope of the juvenile justice system, regarding the adolescents, analysis is made at the level of the criminal/delinquent, which implies offender assessment. Fortunately, there is now consistent scientific evidence regarding personal and social risk factors related to juvenile delinquent conduct that may assist in the proposal of specific public policies. Regarding personal factors, personality is one of the most important explanatory factors for criminal conduct and should be considered. Many investigations indicate that adolescents whose deviant behavior is persistent differ significantly, with regard to some personality traits and adaptive patterns, from adolescents whose deviant behavior is circumstantial (Andrews and Bonta Reference Andrews and Bonta2006; Morizot and Le Blanc Reference Morizot and Le Blanc2003; Soler and López Reference Soler and López2003; van Dam, Janssens, and De Bruyn Reference van Dam, Janssens and De Bruyn2005).
For a better understanding of the concept of personality, it is worth mentioning the specific theoretical orientation adopted. As published by McCrae and Costa (Reference McCrae and Costa1996, Reference McCrae and Costa1997), since birth, babies behave according to some traits/dispositions/temperament to act and feel in one direction. For example, some are more sensitive or more energetic. These characteristics compose a background which comes with the individual and necessarily promote specific interactions in this new person that, once born, is already able to assume an active role in the environment. Therefore, this background both modulates the person’s development as also it is modified during the life-course. Thus, people are born with some propensity but this does not determine personality by itself. Added to this, there is a second component of personality, formed by these specific interactions provided by the individual’s background and by experiences, across the life-course. Also known as adaptive style, is manifested by beliefs, values and attitudes of the individual, about self, others and institutions and psychological defenses against perceptions and feelings. So, the combination of both these levels produces individual differences that mark the way people act and react to the same or similar situations and, over time, people show some degree of stability, consistency and regularity in their behavior.
Thus, in discussing personality, it is considered that it is not intrinsic, immutable or formed disregarding what happens around the individual. In the criminal phenomenon, these beliefs, values and attitudes – which are part of the personality – are identified, in different degrees, with those shared by criminal groups and they increase the probability of committing offenses. That is why personality assessment is one of the crucial ways to understand the misconduct presented by an adolescent. This knowledge can guide programs of prevention, intervention, judicial and socio-educational actions that must be proportional to the characteristics and personal needs of the adolescent (Ridenour, Marchant and Dean Reference Ridenour, Marchant and Dean2001). In this sense, investigations to develop and test instruments for evaluating personality in adolescents, referring to “juvenile delinquency,” have been carried out. A central concern in this area is that the instruments need to focus on relevant aspects of engaging in criminal activity (Andrews and Bonta Reference Andrews and Bonta2006).
In juvenile justice, the superficial understanding of the phenomenon of adolescents, their needs and difficulties and the lack of systematic and well-founded assessments in juvenile delinquency seem to support the maintenance of the reductionist and massive/oppressive practices (Maruschi and Bazon Reference Maruschi and Bazon2014). Concerning psychological assessments, when valid instruments are used, they are generally the same as those used in clinical evaluations, since in Brazil there are no instruments specifically designed to evaluate subjective aspects related to legal issues (Jung Reference Jung2014). Therefore, there are no specific, valid and reliable instruments that can assist in the evaluation of adolescent offenders in the Brazilian context.
Maruschi, Estevão, and Bazon (Reference Maruschi, Estevão and Bazon2013) carried out a study based on three juvenile delinquency assessment instruments – the Youth Level of Service/Case Management Inventory (YLS/CMI), the Jesness Inventory and the Escala Fatorial de Socialização (a Brazilian Instrument based on the Model of the Big Five Personality Factors) – to investigate the psychometric characteristics of the instruments. In addition, the study provided considerations about the auxiliary evaluation of jurisdictional decision in Brazilian juvenile justice, where the study was conducted. The authors pointed out the fact that evaluations in this field suffer from a lack of clear theoretical and methodological guidance. The evaluations are not very systematic and are based on the use of standardized instruments, in Brazil. They refer to others countries that have received academic–scientific as well as governmental investment in a system of evaluation of adolescents in the field of justice, and pointed out that the reliable evaluation of factors that support such conduct seems indispensable for the adequacy of the processes that occur in this scope. Persistent delinquent behavior is supported by contextual and personal variables. Thus, the use of standardized and empirically based instruments, as is done in other countries, can contribute to minimize the discretion that governs the application of judicial/socio-educational measures. The systematic evaluation of adolescents, in this context, can too contribute to the identification of the aspects that are really relevant to public policies related to programs of primary prevention and to treatment of juvenile offenders.
Although the Maruschi et al. (Reference Maruschi, Estevão and Bazon2013) study is recent, this issue had already been recognized many years before, and, to deal with this, since the 1980s a series of studies had been carried out aiming to adapt and validate the Jesness Inventory to the Brazilian sociocultural context (Bertini and Estevão Reference Bertini and Estevão1986; Bertini, Tostes, and Estevão Reference Bertini, Tostes and Estevão1984; Estevão and Bichuette Reference Estevão and Bichuette1985a; Estevão and Bichuette Reference Estevão and Bichuette1985b; Estevão and Stephaneck Reference Estevão and Stephaneck1983; Manzi-Oliveira Reference Manzi-Oliveira2012; Maruschi Reference Maruschi2010; Panosso Reference Panosso2008; Pestana and Bazon Reference Pestana and Bazon2005).
Carl Frandall Jesness, an American behavioral psychologist, developed the Jesness Inventory in the 1960s. Based on his clinical experience, research on programs aimed at juvenile offenders and on the scientific literature available about this theme, Jesness developed a specific instrument for the evaluation of juvenile offenders, which evaluates adaptive patterns (opinions/thoughts, beliefs, attitudes, perceptions/distortions, feelings/emotions, psychological defenses), and some basic dispositions such as impulsivity/sensation seeking and hostility. It results in scores on 10 personality scales, two scales that indicate the possibility of disorders and an index referring to the potential risk of recidivism in an offender’s behaviors (Jesness Reference Jesness.1983, Reference Jesness2003). Investigations into this instrument have already been carried out in its context of origin, demonstrating its power to differentiate groups of adolescent offenders from non-offenders (Jesness Reference Jesness2003; Martin Reference Martin1981; Singh Reference Singh1983), confirming its validity criterion.
So, in the 2000s, it was updated with the inclusion of five new items, totalling 160, and two new scales to screen some indications of psychiatric disorders, based on the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV). Like this, the Jesness Inventory–Revised (JI-R; Jesness Reference Jesness2003) was proposed, consisting of 10 personality scales, two subscales referring to DSM-IV typologies and an index that is a composite measure generated from weightings of selected personality scales and reflects the extent to which the youth engages in delinquent acts and contravenes general social norms. The personality scales are: social maladjustment (SM), value orientation (VO), immaturity (Imm), autism (Au), alienation (Al), manifest aggression (MA), withdrawal–depression (Wd), social anxiety (SA), repression (Rep) and denial (Den); the two subscales referring to DSM-IV typologies are conduct disorder (CD) and oppositional defiant disorder (ODD); the index is the Asocial Index (AI) obtained through the application of a formula that combines scores of eight scales (SM, OV, Au, Al, MA, Wd, SA and Rep). Besides this, the major aspect of the revision was the establishment of an updated normative standard (Jesness Reference Jesness2003). Although the JI-R has been normed for youth as young as age 8 years and adults, the most popular use of the tool has been with adolescents, particularly those demonstrating antisocial, behavioral and emotional concerns.
Studies by Manzi-Oliveira (Reference Manzi-Oliveira2012) in Brazil, Wenger-Amengual (Reference Wenger-Amengual2010) and Ergas and Narváez (Reference Ergas and Narváez2015) in Chile, and Fariña and Andrés-Pueyo (Reference Fariña and Andrés-Pueyo2008) and Fariña and Martinez (Reference Fariña and Martínez2009) in Spain on the JI-R obtained evidence of criterion validity similar to its original context, whereupon some of the scales of the instrument are able to discriminate groups of judicialized and non-judicialized adolescents.
In summary, in different sociocultural contexts, the JI-R has been proved to be an efficient measure to discriminate offenders from non-offenders and, as expected, in view of its theoretical basis and studies by Jesness (Reference Jesness2003), SM and AI are the most prominent scales for this propose. The VO scale also stood out in all referenced studies as an important measure, as well as the MA scale, which discriminated the groups.
Regarding simultaneous validity, which measures the instrument’s power to rank properly offenders and non-offenders, Jesness (Reference Jesness2003) obtained evidence that corresponds to those obtained in the original Jesness Inventory (Jesness 1962). The AI scores were used as the main predictor, controlling the variables age and gender in the analyses. This resulted in 634 correct classifications out of a total of 778 offenders (81.5% accuracy), and in 2,740 correct classifications out of a total of 3,421 non-offenders (80.1% accuracy). Also, when the SM scores were focused, it correctly classified 615 of the 778 offenders (79% accuracy) and 2,596 of 3,421 non-offenders (75% accuracy).
The present study intends to offer a contribution in this direction, opting to continue the efforts already made in order to promote the adaptation and validation of the JI-R to the Brazilian sociocultural context, for its use in assessment of adolescents in conflict with the law. It is necessary to emphasize that adaptation of instruments intended to measure an aimed construct developed in other sociocultural contexts is better than elaborating a totally new instrument. Besides increasing the chances of cross-cultural research, it implies a greater facility for theoretical framework of the instrument, a greater similarity of the evaluation methods, with the possibility of comparing several samples based on its scores, and a lower cost in financial terms and time (Borsa, Damásio, and Bandeira Reference Borsa, Damásio and Bandeira2012; Hambleton Reference Hambleton2005).
OBJECTIVE
The purposes of the present study were to obtain and evaluate new evidence of validity of the Brazilian Jesness Inventory–Revised (JI-R-Br; Manzi-Oliveira Reference Manzi-Oliveira2012):
-
∙ Concurrent validity, in order to verify the power of the JI-R-Br to correctly discriminate adolescent offenders from non-offenders, taking into account the external variable associated with the evaluated construct – criminal engagement – using, in this case, the indicator “whether or not they have been in conflict with the law” (juicialized);
-
∙ Simultaneous validity, to verify the power of the JI-R-Br, specifically the SM and AI scales, to properly classify adolescents into the membership groups (judicialized and non-judicialized groups).
METHODS
Participants
The present study consisted of 848 male judicialized adolescents (in conflict with the law) recruited from probation or correctional units of the Brazilian Juvenile Justice System, whose mean age was 16.9 years (SD=1.2), and 1,475 male non-judicialized adolescents (as a comparative group) recruited from state and private schools, whose mean age was 14.8 years (SD=2).
Materials
The samples were assessed with the Brazilian version (Manzi-Oliveira Reference Manzi-Oliveira2012) of the Jesness Inventory-Revised (JI-R; Jesness Reference Jesness2003), a self-report measure consisting of 160 true/false questions. According to the aforementioned, the inventory has 10 personality scales that measure key individual traits and attitudes (SM, VO, Imm, Au, Al, MA, Wd, SA, Rep and Den), two subscales referring to DSM-IV typologies (CD and ODD) and an index referring to the potential risk of recidivism in offenders’ behaviors (AI). In addition, the JI-R has two validity scales, the Lie scale and Random Response scale, which assess potentially invalid response patterns. The raw score obtained at each scale is transformed into a standardized score taking into account the age of the individual based on a normative reference population. Internal consistencies of the JI-R range from .61 to .93, and a test–retest 1-year term reliability ranges from .50 to .72 (Jesness 2003). In a recent study conducted in Brazil, test–retest reliability ranged from .45 to .71 (Bazon Reference Bazon2016).
Procedure
Trained professionals administered the JI-R-Br to judicialized adolescents, individually, in an exclusive room within the institution where the youth was complying with the judicial measure. In relation to the adolescents from schools, trained professionals administered the inventory collectively, in a room reserved by the school staff for this purpose.
Data Analysis
Regarding concurrent validity, the mean scores obtained for each of the JI-R-Br scales by the judicialized adolescents and the non-judicialized adolescents were compared, by age group, using Student’s t test for independent samples in the cases that the assumptions for parametric methods were fulfilled: the variables to be compared must be in a continuous scale, their distributions must be approximately normal, there must be homogeneity between variances and independence between observations, and there are no outliers. Scales that did not achieve these assumptions had their means compared by the non-parametric Mann–Whitney U test.
Concerning simultaneous validity, assuming that the SM and AI scales are the ones that have the best power to diagnose the problem in focus (delinquent engagement), it was hypothesized that they would be able to correctly classify a high proportion of adolescents taking into account the group to which they belong (judicialized v. non-judicialized). Thus, the frequency of adolescents observed at each scoring of the SM and AI scales and their respective relative and accumulated frequencies were calculated. Then, focusing on the accumulated frequencies, it was possible to identify the cut-off points, for each scale, that would better distinguish the judicialized adolescents from the non-judicialized, that is, that point below which most the adolescents of a group would be, and above it, most of the other.
RESULTS
Regarding the analysis of the evidence of validity of the instrument, as mentioned above, a comparison between judicialized and non-judicialized groups was made using the raw mean scores obtained in each of the JI-R-Br scales by age group, aiming to assess the concurrent validity, as the Table 1 summarizes.
Table 1 Comparison between the average raw scores of the judicialized and non-judicialized adolescents for the different scales of the Brazilian Jesness Inventory–Revised, by age group
Regarding simultaneous validity, the power of the SM and AI scales to correctly classify the adolescents investigated (n=2,328) into their membership groups – judicialized and non-judicial – was verified. Table 2 presents the proportions of non-judicialized and judicialized adolescents distributed in each range corresponding to the raw scores on the SM scale, denoting that the raw score of 31 is the one that best places the judicialized adolescents on one side and the non-judicialized adolescents on the other. About 73% of non-judicialized adolescents score up to 31 points, while 68.4% of the judicialized scored above 31. This means that if a judicialized teen responds to the instrument, the chance that he will score above 31 on the SM scale is approximately 68.4% (95% confidence interval=3.11%), while that of a non-judicialized adolescent is 27.4% (95% confidence interval=2.28%). In addition, up to 10 points the chance of the individual being in the group of judicialized adolescents is the lowest.
Table 2 Relative and cumulative proportions of judicialized and non-judicial adolescents in social maladjustment (SM) scale scores
a The raw score of 31 is the one that best places the judicialized adolescents on the one side and the non-judicialized adolescents on the other.
Table 3 shows the proportions of judicialized and non-judicialized adolescents in each range on the AI measure – based on raw scores. The results indicate that the score of 22 is the one that best places the adolescents judicialized on one side and the non-judicialized on the other. About 74% of the non-judicialized scored as high as 22 in AI whereas about 80% of the judicialized scored above 22. In addition, it can be said that up to 12 points the chance of the individual being in the judicialized group of adolescents is the lowest. This means that if an offender adolescent responds to the instrument, his chance of scoring above 22 on the AI scale is approximately 80.3% (95% confidence interval=2.66%), while that of a non-offender adolescent is 26.2% (95% confidence interval=2.25%).
Table 3 Relative and cumulative proportions of judicialized and non-judicial adolescents in Asocial Index (AI) scale scores
a The raw score of 22 is the one that best places the judicialized adolescents on the one side and the non-judicialized on the other.
DISCUSSION
Regarding the validity analysis of the JI-R-Br, Table 1 shows, in five of the 13 scales, the judicialized and non-judicialized adolescents presented significantly different mean scores at all ages: SM, VO, Al, Rep and AI, in which the judicialized had always higher average scores than non-judicialized, as expected. In Imm and Wd, no significant difference was observed between the adolescents of the two groups only at the age of 13 years. In this case, it is worth noting that the number of juvenile offenders at that age compared to their counterparts was very small (n=5). However, considering that in the other five age comparisons regarding these scales, significant differences between groups were detected; even working with this small “n,” it is possible to consider that the differences in the characteristics represented by both dimensions (Imm and Wd) become more pronounced after the age of 13 years, in the transition from the first to the second stage of adolescence.
In fact, observing the results obtained in other scales, it is noticed that the significant differences between the adolescents of the two groups seem to be concentrated in the higher age groups. Perhaps this is due to the fact that most of the characteristics evaluated by the constructs tend to become less pronounced with development, as Jesness (Reference Jesness.1983, Reference Jesness2003) warns in describing the scales. Thus, it can be thought that the most marked differences between offenders and non-offenders in the older ages would represent the developmental gap experienced by the offenders, which would increase and become more evident with age. This is observed, for example, in the data in Table 1, for the SA and Den scales.
Concerning the CD and ODD scales, it is worth noting that there were no answers to all the items that compose these scales for part of the sample of judicialized adolescents, especially for those aged between 13 and 14 years, which made it impossible to compare these age groups. However, there were significant differences in the dimensions represented by these scales only at the age of 18 years for CD, in which judicialized adolescents scored higher, and at the age of 16 years for ODD, in which non-judicialized adolescents scored, unusually, higher on this. For ODD, it is possible that the non-judicialized were more honest in their responses, at this age. It is possible too that there are some problems with the content of the items that can be interpreted positively by the adolescents, as an assertive response, not as an antisocial response. Interestingly, also, there were no statistically significant differences between the judicialized and non-judicialized adolescents, at any age, on the MA scale. It was expected that this scale would differentiate the groups, like in other studies. In this case, it is possible too that there are some problems with the content of the items. However, it is important to consider that delinquency is not a synonym of aggression. So it is possible too that in the samples analyzed, the aggression levels are the same among judicialized and non-judicialized adolescents.
In all the other JI-R studies referenced in this paper, carried out in different sociocultural contexts, some Jesness scales showed good accuracy in discriminating groups of judicialized and non-judicialized adolescents, as has been verified in the present study. Manzi-Oliveira (Reference Manzi-Oliveira2012), using raw scores, found that the groups in her research presented significant differences in five of the 10 personality scales she analyzed: SM, Imm, Au, AM and Rep and AI (noting that she did not use the scales CD and ODD). No significant differences were found between the groups on the VO, Al, Wd, SA and Den scales. However, the author indicates that she did not control the variable age and verified that it would have interfered with the results of most of the scales; only Rep and Wd would not have been influenced. In order to control this variable, considering its impact on the scores, a new comparison of the groups was carried out, based on their standardized scores and, thus, it was observed that more scales started to differentiate the groups: SM, VO, Au, Al, AM, Rep and IA. With this procedure, there was no significant difference between the groups on the Imm, Wd, SA and Den scales.
In the Chilean context, Wenger-Amengual (Reference Wenger-Amengual2010) points out that there were significant differences in average raw scores obtained from population and offender groups in six of the 12 scales (SM, VO, Au, MA, AI and CD) and did not find significant differences between the groups in Rep and Den. It should be noted, however, that this author did not control the variables age and gender, which may have caused some bias in the results obtained. Likewise, Ergas and Narváez (Reference Ergas and Narváez2015), also in Chile, compared groups of judicialized and non-judicialized adolescents, in order to verify the accuracy of the Jesness scales to discriminate them, without, however, controlling age and gender, which also implies some reservation regarding the results presented, taking into account the importance of these two variables in the evaluation of the characteristics represented in the scales of the inventory (Jesness Reference Jesness2003). Thus, they found that the offenders differed significantly from non-offenders in 10 of the 12 scales: SM, VO, Imm, Au, Al, MA, Rep, Den, CD, ODD and IA. In a complementary method, using Cohen’s d to estimate the power of each scale, the authors argue that all JI-R scales significantly discriminated the groups and that the largest effect sizes were in SM, VO, CD and AI, and the smallest effect sizes were Au, Al and Rep.
In the Spanish context, according to a Fariña and Andrés-Pueyo (Reference Fariña and Andrés-Pueyo2008) study, young offenders scored significantly higher than the control group in eight of the 12 scales: SM, VO, Imm, Au, Al, MA, AI and CD. Also, Fariña and Martinez (Reference Fariña and Martínez2009) found that young offenders were compared to those offenders studied by Fariña and Andrés-Puyeo (2008), showing that the offenders in both studies presented a very similar profile, with high mean scores in SD, VO, Imm, Au, Al, MA, AI, CD and ODD. Lastly, in the Jornet-Gibert et al. (1999) study, also linked to the same Spanish research group, comparing adults with and without a history of traffic offenses related to alcohol intake, it was found that the offenders differed from non-offenders in SM, VO, Imm, Au, Al, MA and Wd.
In summary, it can be said that in different sociocultural contexts, the JI-R has proved to be an efficient measure to separate offenders and non-offenders and, as expected, in view of its theoretical framework and studies by Jesness (Reference Jesness2003), SM and AI are the most prominent scales for this propose The VO scale, in particular, also stood out in all the studies as an important measure, as well as the MA scale, which discriminated the groups in all the studies, except for the present one, which was not expected. In fact, the result regarding MA, as well as that regarding ODD, obtained in the present study, contradicting in a more forceful way the expectations created in the light of the theory underlying the JI and the definitions of such scales, will deserve a closer look, taking into account, perhaps, the answers to each of the items that compose these scales, in order to raise hypotheses about what occurred.
Concerning simultaneous validity, referring to the power of the inventory to correctly classify individuals in terms of belonging to a group, based on the criterion of “being or not judicialized,” as an indicator of whether or not they have any delinquent engagement (as a function of psychological characteristics), it can be said that the scales tested in this sense – SM and AI – presented very satisfactory performances. Focusing on SM, it was found that more than 70% of the non-judicialized adolescents scored up to 31, while more than 68% of the judicialized adolescents scored above 31. Focusing on AI, it was verified that more than 70% of the non-judicialized adolescents scored up to 22, while about 80% of the judicialized scored above 22. That is, in both scales, such cut-offs are able to adequately discriminate large proportions of individuals; in terms of the real group of belonging, it can be said that the correct classification was far greater than chance.
In the studies implemented by Jesness (Reference Jesness2003), it was verified that the SM scale was able to correctly classify 75% of the non-judicialized and 79% of the judicialized adolescents, while AI was able to correctly classify 80.1% of the non-judicialized and 81.5% of the judicialized. In general, these rates and those found in the present study are not very different, specifically with regard to AI, which is very important considering that this scale was designed for diagnostic purposes, measuring the antisocial motivation of the individual that represents an important psychological aspect that cannot be overlooked in the psychosocial assessment and interventions in this field. Thus, it can be said that the instrument’s performance to correctly classify adolescents judicialized in the Brazilian context is practically the same as in the context of origin.
Thus, concerning these validity data and those that have occurred in other sociocultural contexts and in the context in which it was elaborated, it can be said that JI-R-Br presents sufficient evidence that it evaluates the constructs for which it was proposed. In agreement with Fariña and Andrés-Pueyo (Reference Fariña and Andrés-Pueyo2008), this is an instrument capable of making an important contribution in the field of juvenile justice, especially considering that there is no other instrument as specific as the JI-R. Its relevance in this field is also due to the few resources to carry out systematic evaluations that aid understanding of adolescents’ psychological functioning and behavior, as well as evaluating the quality of the programs offered to adolescents based on the improvement in psychological functioning.
Therefore, the validity evidence of the JI-R-Br presented here helps to stablish an instrument that can improve the assessment practices of the adolescents who are in the justice system and, in this continuum, the services offered to them. Besides that, these outcomes, added to others from previous research (Bertini and Estevão Reference Bertini and Estevão1986; Bertini et al. Reference Bertini, Tostes and Estevão1984; Estevão and Bichuette Reference Estevão and Bichuette1985a; Estevão and Bichuette Reference Estevão and Bichuette1985b; Estevão and Stephaneck Reference Estevão and Stephaneck1983; Manzi-Oliveira Reference Manzi-Oliveira2012; Maruschi Reference Maruschi2010; Panosso Reference Panosso2008; Pestana and Bazon Reference Pestana and Bazon2005), answer a demand placed by Maruschi et al. (Reference Maruschi, Estevão and Bazon2013) for obtaining standardized and empirically based instruments, as is done in other countries where government initiatives appropriate the scientific work in the area to elaborate specific public policies for the issue of adolescents in conflict with the law. Thus, this study may contribute to Brazilian research towards achieving appropriate understanding and management in what concerns the adolescent offender – psychological variables, in this case.
Future studies – although these results are very positive – need to have better indicators of the offender behavior to be used as the dependent variable in the analysis in question. The fact that the adolescent has been in conflict with the law (judicialized), although it is an admissible criterion, is an inaccurate criterion regarding the phenomenon in focus – that, is the delinquent engagement – more accurately measured, in the JI, antisocial orientation (SM and AI). Thus, in future investigations, one can think that a more refined criterion for studies aiming to obtain evidence of concurrent criterion validity and simultaneous validity, when dealing with official data, is the number of passes of the adolescent through the justice system, for example. Another way will be to search for unofficial data, that is, to deal with self-reported data related to the involvement of adolescents in the practice of crimes, in which case, it is possible to work with only samples of adolescents from the population. In addition, in the future, new research to obtain evidence of predictive and construct validity should be carried out.
Rafaelle C. S. Costa is a psychologist at the University of São Paulo and is a member of the Studies and Research in Development and Psychosocial Intervention Group (GEPDIP; University of São Paulo).
André V. Komatsu is a psychologist at the University of São Paulo and is a Master of Science. He is currently a PhD student at the University of São Paulo. He is a member of the Studies and Research in Development and Psychosocial Intervention Group (GEPDIP; University of São Paulo).
Marina R. Bazon, PhD is currently a Professor at the University of São Paulo. She is the coordinator of the Studies and Research in Development and Psychosocial Intervention Group (GEPDIP; University of São Paulo).
