Research design

An empirical, quantitative, descriptive, ex post facto, cross-sectional study, using a survey and a non-probabilistic sample (^{Montero & León, 2007}).

Participants

The participants in this study were 191 psychologists practicing in any Latin American country who voluntarily responded anonymously to an Internet survey, 76.4 % of whom were women (n = 146). The mean age was 42.08 years (SD= 9.34), ranging from 23 to 74 years with no significant differences in age by sex (t(61.51) = -0.46, p= .650). Professionals from 10 Latin American countries participated in the survey. For comparative analysis purposes, the countries were grouped into three regions: Argentina was kept as a single region because it was the country with the highest number of participants. The remaining countries were grouped as follows: Rest of South America (in brackets the number of participants per country) (Venezuela (1), Uruguay (18), Chile (11), Colombia (6), Peru (5), Ecuador (5)); and Central America (Guatemala (1), Costa Rica (28), Mexico (11)).

Procedure and materials

The survey was designed in the Google Forms platform and consisted of direct closed-answer questions grouped into five sections: a) Sociodemographic characteristics (sex, age, country), b) General academic training (profession, educational level, year, and university where degree was obtained), c) Forensic-specific training (type of training, type of institution, year of that training, and year of last update), d) Forensic professional practice (field, hours per week, years, current country, minimum requirements), e) Psychological assessment areas: For each area (mental state at the time of the crime, risk of general violence, risk of sexual violence, simulation and psychological harm), we asked about the number of assessments in the last year and the frequency of use of a series of techniques selected ad hoc for each area (from 0 = never to 4 = very frequently). The techniques were proposed along with examples, as follows: interview, Rorschach test, projective tests (e.g., TAT, HTP), multiscale personality inventories (e.g., MMPI, PAI, MCMI), neuropsychological tests (e.g., Barcelona Test, Bender Test, Wisconsin Test), cognitive and ability assessment techniques (e.g., Wechsler scales), actuarial instruments for violence risk assessment (e.g., VRAG, COVR), structured professional judgment guides for violence risk assessment (e.g. HCR-20, SVR-20), psychopathy assessment checklists (e.g. PCL-R, PCL:SV), clinical psychopathological assessment scales (e.g. SCL-90, BDI, STAI), specific techniques for the assessment of simulation (e.g., SIRS, TOMM), post-traumatic stress disorder assessment instruments (e.g., PSS scales, CAPS-I, SPRINT), and others.

Initially, a pilot test was conducted with three professionals who analyzed aspects of comprehensibility or fatigue that could affect the reliability or validity of the responses (^{Bhattacherjee, 2012}). After the adjustments suggested, the survey was disseminated via email and on social networks (Facebook, WhatsApp, LinkedIn and Instagram) from April 23 to August 28, 2021. Periodic reminders were sent on the three Fridays following the launch, forwarding an e-mail, and making a new publication in the networks (see ^{Sánchez Fernández et al., 2009}). It was not possible to control the number of people who had actual contact with the survey because, due to ethical considerations, no confirmation of message receipt or user registration was requested. Respondents could complete the survey only once, all questions required responses to proceed with the survey, a ‘not applicable’ response option or similar was always provided. The survey administration followed the recommendations of the Checklist for Reporting Results of Internet E-Surveys protocol (^{Eysenbach, 2004}).

Ethical considerations

The protocol of this study was evaluated by the Ethics Committee of the Instituto de Investigaciones Psicológicas of the Universidad Nacional de Córdoba with a subsequent favorable approval on December 11, 2020. The survey commenced with an information sheet, followed by the informed consent form.

Data analysis

The data were analyzed with the Statistical Package for the Social Sciences (SPSS), version 22. Analyses of absolute and relative frequency distributions were conducted, and the results presented in contingency tables. Means and standard deviations were calculated for interval variables. To assess whether the variables showed variations in their distribution across regions, comparisons were conducted using contingency tables, and the Chi-Square statistic was calculated to test the hypothesis of independence between groups.

Academic training and professional practice

Table 1 presents the academic characteristics of the professionals. Approximately half (n = 90) indicated having completed specific and accredited training in legal psychology or a related area. Of these, 64 did so at a university and the rest at an association or institution other than a university, generally professional associations or training schools attached to legal or police institutions. Another 47 were undergoing specific and accredited training at the time of the survey, 41 at a university and 6 at different institutions. There was a significant negative correlation between age and the year in which specific and accredited training was completed (r = - .36, p < .001), indicating that the younger professionals tended to have more complete and up-to-date training.

Table 1: Academic characteristics of participants 

^Notes:a Less than 6 months. ^b 1 to 2 years.

Table 2 presents aspects of professional practice. Seventy-five percent of respondents had five or more years of practice and reported different requirements for conducting psychological assessments in the legal settings in which they practiced. The minimum reported requirement was associated with the practitioner’s region (x ²(4) = 13.80, p= .008). Some 17.5 % of practitioners in Central America reported that a specialty or higher level was required, while that proportion was 15.2 % in the rest of South America and 7.7 % in Argentina. Notably, in no case did all the participants from the same country or region agree on the minimum requirement. On the contrary, their responses were dispersed across the different categories. Those with a higher level of education tended to report a more restrictive minimum level requirement (i.e., a higher academic degree) than those with a lower academic degree (c ²(8) = 32.19, p< .001).

Table 2: Professional practice in Forensic Mental Health: years, country and requirements 

^Note:a In order to practice your professional activity in legal contexts in your country, what is the minimum academic requirement?

Assessment areas

State of mind at the time of the act or imputability

Among those who conducted assessments in this area in the past year (n = 130, 68.1 %), the interview was the most frequently used technique followed by personality inventories (Table 3). There were significant variations in the frequency of use of techniques across regions, with a tendency in Argentina and South America to use projective tests, and in Central America to use structured instruments (e.g., personality inventories, cognitive assessment tests, ability tests, clinical scales). The footnotes to each table detail the differences found between regions.

Table 3: Assessment of mental status or non-imputability: Use of techniques in the past year (n = 130) 

^Notes: SPJ: Structured Professional Judgment. The asterisks indicate significant differences according to regions at the level of *p < .05 **p < .01 ***p < .001. ^a More frequent use in Argentina, followed by South America. ^b More frequent use in Central America, followed by South America. ^c More frequent use in Central America, followed by Argentina. The "Very often" category is the reference.

Risk of general violence

More than two-thirds of professionals (n = 138, 72.3 %) had conducted at least one general violence risk assessment in the past year. Again, the most frequently employed techniques were the interview, followed by personality inventories (Table 4). Structured professional judgment (SPJ) guides for violence risk assessment are specific instruments for this task, but their use was reported less frequently than other non-specific instruments such as projective tests and clinical scales. Variations by region were similar to those reported above, with a preference in Argentina and South America for projective techniques and the Rorschach in these assessments, in contrast to Central America where psychometric tests were preferred. However, there were no differences between regions in the frequency of use of specific techniques for violence risk assessment (actuarial or SPJ).

Table 4: General violence risk assessment: Use of techniques in the last year (n= 138) 

^Notes: SPJ: Structured Professional Judgment. Asterisks indicate significant differences by region at the *p < .05 **p< .01 ***p< .001 level. ^a More frequent use in Argentina, followed by South America. ^b More frequent use in Central America, followed by South America. ^c More frequent use in Central America, followed by Argentina. The "Very often" category is the reference.

Risk of sexual violence

Two-thirds of the participants indicated having conducted risk assessment for sexual violence in the past year (n = 124, 65.8 %). As in the case of general violence, interview and personality inventories were used more frequently here (Table 5). Specific instruments (SPJ guides and actuarial instruments) were used less frequently than non-specific instruments, such as projective tests and clinical scales. Regarding regional differences, the same trend was observed toward the use of projective techniques and Rorschach in Argentina and South America, and psychometric tests in Central America.

Table 5: Sexual violence risk assessment: Use of techniques in the past year (n = 124) 

^Notes: SPJ: Structured Professional Judgment. Asterisks indicate significant differences by region at the *p <.05 **p< .01 ***p< .001 level. ^a More frequent use in Argentina, followed by South America. ^b More frequent use in Central America, followed by South America. ^c More frequent use in Central America, followed by Argentina. The "Very often" category is the reference.

Simulation

Slightly more than half of the respondents conducted at least one such assessment in the past year (n = 99; 51.8 %). The results were similar to those reported for other areas of assessment. That is, the most commonly employed techniques were the interview, followed by personality inventories (Table 6). Specific techniques (e.g., TOMM) were used less frequently than non-specific instruments, such as clinical scales and projective tests. Finally, the differences between regions follow the trend reported in the previous sections, with the exception of neuropsychological tests, which are reported, for the first time, to be used more frequently in Argentina.

Table 6: Simulation Assessment: Use of techniques in the past year (n = 99) 

^Notes: SPJ: Structured Professional Judgment. Asterisks indicate significant differences by region at the *p < .05 **p < .01 ***p < .001 level. ^a More frequent use in Argentina, followed by South America. ^b More frequent use in Central America, followed by South America. ^c More frequent use in Central America, followed by Argentina. d More frequent use in Argentina, followed by Central America. The “Very often" category is the reference.

Psychological harm

A total of 138 professionals (72.3 %) assessed psychological harm in the last year. The data show the three findings reported in previous tables (Table 7): a) a predominant use of the interview, followed by personality inventories, b) a more frequent use of non-specific techniques (e.g. clinical scales and projective tests) than specific techniques (in this case, PTSD assessment tests), and finally, c) a preference for projective techniques and the Rorschach in Argentina, compared to Central America where psychometric tests are more frequently used.

Table 7: Psychological harm assessment: Use of techniques in the last year (n= 138) 

^Notes: Asterisks indicate significant differences by region at *p< .05 **p< .01 ***p< .001. ^a More frequent use in Argentina, followed by South America. ^b More frequent use in Central America, followed by South America. The "Very often" category is the reference.

Specific academic training and professional practice

The absence of specific accredited training in legal psychology (or related areas) in approximately a quarter of the participants indicates that accreditation of specific knowledge for professional practice in this field is not a widespread requirement in the region (^{Quintero et al., 2020}). However, exclusive reliance on undergraduate training may carry its risks given that it will tend to be generalist, rather than specific (^{Benito, 2009}). In fact, only a small proportion of professionals indicated having obtained specific training in legal psychology or related areas in their undergraduate training. This may be related to the fact that until recently there was a relative lack of specific postgraduate offerings in this region (^{Barboni & Bonilla, 2022}; ^{Quintero et al., 2020}).

The striking reactivation of Legal and Forensic Psychology in the United States and Europe around the 1960s, strongly reinforced since the 1980s, promoted the proliferation of professional associations, academic societies, conferences and journal publications, fundamentally regulated by principles of scientificity (^{Fariña et al., 2005}; ^{Heilbrun, DeMatteo, et al., 2008}). This phenomenon does not seem to have been replicated with the same intensity generally in Latin America (see ^{Goldstein et al., 2015}).

There are indications that those practitioners who rely more on clinical judgment will be less likely to use instruments, while the reverse is true for those who prioritize evidence-based practices and standardization of assessment processes (^{Neal & Grisso, 2014b}). The preference for clinical judgment in decision-making, rather than an evidence-based model (see ^{Ægisdóttir et al., 2006}), goes back to the direct antecedent of the Daubert criteria, known as the Frye criteria (^{Frye v. U.S., 1923}), whereby the admissibility of a test was established on the basis only of the general acceptance of the test in its disciplinary field. Progressively it has been shown that this single criterion is not sufficient and others were added to strengthen the scientific basis of the evidence. In recent years, there has been a marked development of guidelines for good practice in psychological assessment for experts in legal contexts (^{Dror & Murrie, 2018}; ^{Neal et al., 2022}).

Our results suggest that the incorporation of scientific criteria in psychological assessment in legal settings is an ongoing process in this region, given that a greater proportion of the youngest reported having specific training. This may indicate a trend in Latin America towards accreditable professional specialization and the regulation of the necessary requirements for professional practice in this field. In this context, good practice guidelines play an essential role, and, in the Spanish-speaking world, Spain is one example of their development and application (e.g., ^{APF, 2018}; ^{COPC et al., 2016}).

Despite the limitations found in the survey in the specific training of the professionals, when it was present, it was quite recent (on average, it had occurred in the previous 6 years), which suggests that the professionals, especially the younger ones, are quite up-to-date. However, this finding still raises the question of the curricular contents of the professional training offered, and to what extent these contemplate the scientific standards that the discipline has reached elsewhere.

The answers to the question on the minimum requirements for professional practice in legal areas in each country varied, even among professionals from the same country. This points to the need to establish, communicate, and precisely regulate the minimum requirements for professional practice in this area of great disciplinary specificity. Because of this specificity and its ethical implications, professionals should know this information unequivocally (^{APA, 2013}). Otherwise, there will remain biased practices and beliefs, as indicated by the finding that those with a higher educational level (e.g., master's or doctorate) tended to report higher requirements for professional practice.

Psychological assessment practices in legal contexts

With greater or lesser levels of specialized training, between 40 % and 60 % of the respondents carried out assessments in the areas analyzed. The most frequent were assessments of the risk of general violence and psychological harm, followed by assessments of mental state at the time of the event or non-imputability. Virtually all respondents reported using the interview as an assessment tool in all areas. Given that we did not ask whether the interviews were open-ended or semi-structured, it would be of interest in future studies to explore the level of structure used in these interviews and how it is established. This is essential because, in general, the instruments proposed in this survey are not accompanied by an interview guide, nor is accreditation for their use required in a binding manner. This may cast doubt on the reliability of the data collected in the assessments, and could also explain the lack of evaluator qualification reported in some studies (^{Ireland, 2012}; ^{Muñoz Vicente & López Ossorio, 2016}).

In general, the most frequently employed techniques in the assessment areas explored here were the same. That is, the interview was clearly in first place, followed by multiscale personality inventories, projective techniques, and clinical scales, respectively. Only in the assessment of mental status did neuropsychological techniques rank above clinical scales. Overall, the results indicated the use in legal settings of techniques originally designed for clinical settings. Although the differences between these settings and their implications have been historically pointed out (^{Echeburúa et al., 2011}; ^{Greenberg & Shuman, 1997}; ^{Heilbrun, 1992}), the choice of instruments reported in this survey indicates a limited use in Latin America of specific techniques for the legal field. For example, in the assessment of the risk of violence, both general and sexual, there was a preference for the use of personality inventories, projective tests and clinical scales, rather than the actuarial instruments and structured professional judgment guides that are specific to these assessments. These practices may be based on theoretical frameworks with predominantly classical psychiatric or psychopathological conceptions of dangerousness, considering violence as an emergent of a mental alteration or a particular personality trait (^{Andrés Pueyo & Arbach, 2014}). A previous study indicated this (^{Arbach et al., 2017}), and what is found here seems to show that it has not substantially changed. This occurs despite the evidence on the good psychometric properties shown by specific violence risk assessment tools in Latin American population samples (Arbach & ^{Folino, 2021}; ^{E. Folino & Arbach, 2021}; ^{J. Folino, 2015}).

The preference for generic, rather than specific, tools was also observed in the simulation assessment. That is, the practitioners do not choose those instruments specifically designed for the task and the context. In the case of the assessment of harm, the results are somewhat more encouraging as a higher proportion of professionals indicated often using tools to assess post-traumatic stress disorder. However, even in this case, clinical (generic) scales are used more frequently than specific instruments. This contrasts with the recommendations of good psychological practices that have been generated for legal settings (^{Heilbrun, DeMatteo, et al., 2008}).

Despite the common factors described in the frequency of instrument use, variations were observed according to region, which showed a similar pattern among the different areas of assessment. Projective techniques and the Rorschach were most frequently reported by professionals from Argentina, followed by South America. In contrast, professionals from Central America reported using psychometric tests more frequently. This suggests the influence of a psychodynamic perspective in the former (^{Klappenbach, 2015}), and of evidence-based practice in the latter. This could be due to both historical reasons related to the training of psychologists in the different countries of the region, and geographical reasons, as the countries of northern Latin America showed assessment practices similar to those reported in previous surveys by professionals from North American and European countries (e.g., ^{Archer et al., 2006}; ^{Lally, 2003}; ^{Neal & Grisso, 2014b}).

Limitations and future lines

This study uses non-probabilistic sampling and, therefore, it is not possible to generalize its results beyond the study population (^{Eysenbach, 2004}). Nevertheless, since this is the first survey in which professionals from so many countries in the region participated, and since its data are consistent with previous findings, our results may be useful in the generation of hypotheses whose contrast will allow us to improve the understanding of assessment practices in the region and to adapt them, as necessary, to current scientific knowledge.

At the same time, the study does not escape certain biases inherent to surveys conducted on the Internet. One of these is related to biases in the representativeness of the sample. By using lists of personal contacts, the majority of the sample are professionals geographically close to the authors. In fact, seven Latin American countries are not represented in this survey, and it would be advisable to include them in future studies to favor generalization from cross-national comparisons. The self-selection bias could also have influenced participation by increasing the representation of people with concerns about the topic. Conversely, the self-selective effect may have been counteracted by the participation of people who, although they did not strictly meet the inclusion criteria, still decided to respond to the survey. It is difficult to know how this may have impacted the results, and we believe this is an interesting topic for future exploration.