Personality Dimensions Involved in the Adaptation to the Prison Environment: Evidence from Romanian Violent Offenders

Abstract

Background: Personality traits, particularly those belonging to the Dark Triad (Machiavellianism, narcissism, and psychopathy), may influence violent behavior and adaptation to the prison environment. Methods: The study included 250 male inmates from Bucharest-Jilava Penitentiary, aged between 20 and 67 years. The Emotional Stability Scale (IPIP), the Buss–Perry Aggression Questionnaire, and the Short Dark Triad Scale (SD3) were administered. Statistical analyses included Pearson correlations, multiple linear regressions, and binary logistic regressions. Results: Emotional stability was negatively correlated with aggression (r = −0.48, p < 0.01). Psychopathy significantly predicted physical aggression (t = 11.96, p < 0.001) and anger (t = 9.53, p < 0.001), whereas Machiavellianism predicted verbal aggression (t = 3.14, p = 0.002) and hostility (t = 4.73, p < 0.001). Higher levels of physical aggression were associated with a lower likelihood of participation in educational activities (Exp(B) = 0.93, p = 0.032). Conclusions: The influence of Dark Triad traits on aggression is differentiated, with psychopathy exerting the strongest effect. Low emotional stability increases vulnerability to hostile behaviors. These findings support the need for multidimensional psychological assessments and tailored intervention programs designed to enhance violent offenders’ engagement in educational activities aimed at impulse control and empathy development within the prison environment.

1. Introduction

Numerous studies (Paulhus and Williams 2002; Rauthmann and Will 2011; Muris et al. 2017) conducted on the general population have shown that high scores on dimensions such as aggression, Machiavellianism, narcissism, and psychopathy are associated with difficulties in social and occupational adjustment. For example, aggression has been linked to problems in emotional regulation, hostility, and impaired social integration (Bettencourt et al. 2006).

Similarly, Dark Triad traits have been associated with manipulative behaviors and dysfunctional interpersonal relationships (Paulhus and Williams 2002; Muris et al. 2017). Rauthmann and Will (2011) demonstrated that manipulative tendencies may impair social adjustment, while Jonason and Krause (2013) identified associations between psychopathy, impulsivity, and antisocial behavior. Likewise, Foster and Campbell (2007) observed that elevated narcissism may lead to conflictual relationships and deficient interpersonal cooperation.

The Dark Triad was conceptualized by Paulhus and Williams (2002) as a constellation of personality traits—Machiavellianism, psychopathy, and narcissism—that contribute to dysfunctional behavior patterns. Machiavellianism involves interpersonal manipulation and a lack of morality (Wastell and Booth 2003), psychopathy reflects tendencies toward exploitation and emotional insensitivity (Lee and Ashton 2005), while narcissism is characterized by an excessive need for admiration and a sense of superiority (Foster and Campbell 2007). Although these traits have been extensively studied, their relationship with emotional stability and context-specific adaptation remains controversial.

Emotional stability has been shown to represent an important protective factor. Low levels of emotional stability have been associated with violent tendencies and difficulties in emotional regulation (Tharshini et al. 2021). However, findings across studies are inconsistent: some emphasize emotional stability as a key factor in preventing aggressive behavior (DeLisi et al. 2010), whereas others suggest that its impact depends on the socio-institutional context (Rogers 2019).

Research has extended the investigation of these variables to incarcerated populations. Tharshini et al. (2021) showed that impulsivity and psychopathy are predictors of recidivism and instrumental violence, while Lainidi et al. (2022) highlighted associations between Machiavellianism, psychopathy, and exploitative behaviors. Nevertheless, results regarding the specific role of each trait in prison adjustment remain inconclusive. Studies focusing on prison adjustment (Dellazizzo et al. 2018; Naidoo 2024) emphasize the role of emotional stability in inmates’ well-being but do not clarify how this dimension interacts with Dark Triad traits.

Several studies (Mehrabian 1997; Tat’yana et al. 2016; Rogers 2019) confirm that emotional stability influences adaptive capacity; however, the psychological mechanisms underlying this relationship remain insufficiently explained. In the prison environment, individuals with low emotional stability are at greater risk of hostile behaviors and social isolation. However, few studies have simultaneously examined emotional stability, aggression, and Dark Triad traits within a specific cultural framework, such as the Romanian context.

A further justification for examining this prison population lies in the significant social impact of violent offenses and their important presence within correctional systems. Although the overall rates of violent crime in Europe are lower than in many other regions of the world, these offenses remain a major concern for criminal justice systems due to their severity and social consequences. According to data reported by the United Nations Office on Drugs and Crime (United Nations Office on Drugs and Crime 2023), a total of 3862 intentional homicides were recorded by police authorities across European countries in 2022, representing a 4.4% increase compared to the previous year and observed in 14 of the 27 reporting countries. The same source also documented an increase of approximately 10% in sexual violence offenses in Europe in 2022, with police authorities registering 231,456 such criminal acts.

In Romania, empirical investigations examining the relationship between personality traits and adjustment to incarceration remain scarce. Data published by the National Administration of Penitentiaries (NAP 2024) indicate that more than half of incarcerated individuals are serving sentences for offenses involving physical or sexual violence. Furthermore, the National Strategy for Public Order and Safety 2023–2027 (Government of Romania 2023) reports that, between 2021 and 2023, offenses involving bodily harm or other acts of violence accounted for over 55% of crimes against the person, while sexual offenses increased by approximately 15%.

Bucharest–Jilava Penitentiary, where the present study was conducted, is one of the largest detention facilities in Romania, operating under semi-open and closed regimes. This institution was selected due to its large population of inmates convicted of violent offenses, offering a diverse and relevant sample for analyzing personality traits and mechanisms of prison adjustment.

From a reintegration perspective, this category of inmates poses substantial challenges in terms of engagement in educational, vocational training, and psychological counseling activities, largely due to elevated aggression and emotional instability. Romanian research addressing these issues remains scarce. Săbăreanu and Gonța (2022) noted that the inherently restrictive and tense prison environment may amplify aggressive manifestations, while Aiftincăi and Constantin (2023) emphasized that a carceral climate dominated by hostility and competition intensifies aggressive behavior.

Within this context, the present study aims to examine the role of personality dimensions—specifically emotional stability, aggression, and Dark Triad traits—in inmates’ adaptation to the prison environment. In particular, the study investigates how these psychological characteristics are associated with indicators of institutional adjustment, including levels of aggression and participation in social reintegration activities. By addressing these relationships in a sample of Romanian inmates convicted of violent offenses, the study provides empirical evidence relevant to outlining the psychological profile of violent offenders and to informing the development of targeted intervention strategies tailored to the Romanian prison environment.

2. Materials and Methods

2.1. Research Design and Participants

The present study employed a quantitative, cross-sectional design aimed at examining the relationships between personality dimensions and adaptation to the prison environment. The variables were analyzed based on data collected through standardized self-report questionnaires.

Data collection was conducted between May 2024 and January 2025 and involved a sample of 250 male inmates incarcerated at Bucharest–Jilava Penitentiary. Participants were selected from individuals serving custodial sentences for violent offenses and were aged between 20 and 67 years. Participation in the study was voluntary, and only inmates who provided informed consent were included in the final sample.

Within the Romanian penitentiary system, inmates may receive institutional credits for participation in educational, rehabilitative, or research-related activities. In accordance with this institutional framework, inmates who participated in the present study received three institutional credits. These credits do not represent direct monetary compensation, but form part of the institutional system used to monitor inmates’ engagement in reintegration activities. Accumulated credits may contribute to recommendations for certain rewards in accordance with legal regulations (e.g., additional visits or communication opportunities with family members, supplementary packages, or temporary leave from the penitentiary).

2.2. Procedure

Data collection took place within the penitentiary under conditions consistent with institutional regulations governing research involving people deprived of liberty. Participants were informed about the purpose of the study, the voluntary nature of participation, and the confidentiality of their responses. They were also informed that their decision to participate or decline would have no influence on their institutional status, disciplinary record, or conditions of detention.

The questionnaires were administered in person by a psychologist working within the penitentiary institution. Participants received standardized instructions regarding the purpose of the study, the voluntary nature of participation, and the confidentiality of their responses. They completed the questionnaires individually in a supervised setting within the penitentiary. Standardized instructions were provided prior to administration, and the researcher was present to clarify any procedural questions, without influencing participants’ responses.

Participation in the study complied with the ethical principles governing research involving human participants and with the institutional regulations applicable within the Romanian penitentiary system.

In the present study, adaptation to the prison environment is conceptualized as inmates’ capacity to adjust to institutional rules and to engage constructively in activities that support social reintegration. Consistent with previous research on institutional adjustment, adaptation was operationalized through two types of indicators: behavioral indicators and participation in reintegration activities. Specifically, levels of aggression were considered indicators of maladaptive adjustment, while participation in educational, vocational training, cultural, moral-religious, and psychological counseling activities was treated as an indicator of constructive engagement within the institutional environment.

Emotional stability was not treated as a direct indicator of adaptation in the present study. Rather, it was analyzed as a personality trait reflecting individuals’ capacity for emotional regulation and stress management, which may influence behavioral indicators of institutional adjustment such as aggression and participation in reintegration activities.

2.3. Instruments

Three standardized psychological instruments were used to assess the variables included in the study. These instruments were administered in their Romanian-language versions, available through the ResearchCentral platform, a research initiative developed by TestCentral and the Assessment and Individual Differences Lab (University of Bucharest), aimed at providing Romanian researchers with psychological instruments translated or validated on the Romanian population.

Emotional stability was assessed using the Emotional Stability Scale from the International Personality Item Pool (IPIP; Goldberg et al. 2006), consisting of 10 items measuring emotional balance, self-control, and resistance to stress. The scale was drawn from the Romanian adaptation of the International Personality Item Pool (Iliescu et al. 2015). Responses were recorded on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). A sample item is: “I remain calm even in stressful situations.” Aggression was measured using the Buss–Perry Aggression Questionnaire (Buss and Perry 1992), which assesses four dimensions of aggression: physical aggression, verbal aggression, anger, and hostility. The instrument comprises 29 items rated on a five-point Likert scale (1 = extremely uncharacteristic of me to 5 = extremely characteristic of me). An example item is: “When provoked, I may react violently.” Dark Triad traits—Machiavellianism, narcissism, and psychopathy—were assessed using the Short Dark Triad Scale (SD3; Jones and Paulhus 2014). The scale includes 27 items rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). A representative item is: “I am willing to manipulate others to achieve my goals.”

The Emotional Stability Scale was administered in its Romanian version, which underwent a level 2 cultural validation process to ensure the adequacy of its concepts and items for the Romanian population. The other psychological instruments used in the study were also administered in Romanian, each having undergone a level 1 cultural validation.

In addition, an omnibus questionnaire was administered to collect socio-demographic information (age, educational level, marital status, number of prior convictions) and data regarding inmates’ participation in social reintegration activities. The questionnaire also collected information regarding disciplinary rewards and sanctions. These variables were recorded as contextual information; however, they were not included in the analytical models because many of the reported events had occurred in previous detention contexts or institutional settings and were therefore not directly comparable within the current institutional environment.

2.4. Hypotheses

Based on the theoretical framework and previous empirical findings, the following hypotheses were formulated:

H1. 

Significant negative correlations exist between aggression and emotional stability among inmates convicted of violent physical offenses.

H2. 

Among inmates convicted of violent offenses, higher levels of aggression negatively predict participation in social reintegration activities.

H3. 

Dark Triad personality traits—Machiavellianism, narcissism, and psychopathy—differentially influence levels of aggression among inmates convicted of violent offenses.

3. Data Analysis

Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 20.0 (IBM Corp 2011, Armonk, NY, USA). Descriptive statistics (mean, standard deviation, skewness, and kurtosis) were computed for all variables, and internal consistency was assessed using Cronbach’s alpha coefficients.

To examine the relationships between variables, Pearson correlation coefficients (r) were calculated. The magnitude of the correlations was interpreted following the guidelines proposed by Gignac and Szodorai (2016), where values between 0.10 and 0.19 indicate small effects, values between 0.20 and 0.29 indicate moderate effects, values between 0.30 and 0.39 indicate large effects, and values of 0.40 or higher indicate very large effects.

For the multiple linear regression analyses, the magnitude of the effects was interpreted based on the adjusted coefficient of determination (R²adj), following recommendations discussed in Funder and Ozer (2019) and Gignac and Szodorai (2016). According to these guidelines, values between 0.04 and 0.09 indicate small effects, values between 0.10 and 0.15 indicate moderate effects, values between 0.16 and 0.25 indicate large effects, and values above 0.25 indicate very large effects.

For binary logistic regression models, effect size was interpreted using the Nagelkerke pseudo-R² index. Following the interpretative framework proposed by Funder and Ozer (2019), values between 0.02 and 0.09 indicate small effects, values between 0.10 and 0.24 indicate moderate effects, values between 0.25 and 0.39 indicate large effects, and values of 0.40 or higher indicate very large effects.

Hypothesis testing was conducted using the following analytical procedures:

-

Pearson correlation analyses to examine the relationships between emotional stability and aggression;

-

binary logistic regression analyses to predict participation in social reintegration activities;

-

multiple linear regression analyses to assess the contribution of Dark Triad traits to different forms of aggression.

4. Ethics and Informed Consent

The conduct of this study complied with all legal requirements set forth by the National Administration of Penitentiaries in Romania for scientific research in penitentiary settings. Data collection was carried out using a paper-and-pencil administration format, conducted either in housing units or in spaces specifically designated for social reintegration procedures within the penitentiary.

The study was conducted in accordance with international and national ethical standards for research involving vulnerable populations. Ethical approval was obtained from the Ethics Committee of the Romanian Academy (approval no. 367/14 December 2023), as well as authorization from the National Administration of Penitentiaries. Written informed consent was obtained from all participants, who were informed that participation was voluntary and anonymous, in compliance with the European General Data Protection Regulation (GDPR).

5. Results

The statistical analyses were conducted using two analytical samples. Pearson correlation analyses and binary logistic regression models were performed on the subsample of inmates convicted of physical violent offenses (n = 180), to examine the relationships between emotional stability, aggression, and participation in reintegration activities specifically within this subgroup. In contrast, the multiple linear regression analyses examining the association between Dark Triad traits and the components of aggression were conducted on the full sample of participants (N = 250).

The sample consisted of 250 male inmates, aged between 20 and 67 years, who were serving final sentences for offenses involving physical or sexual violence. The mean age of the sample was 39 years, and the average educational level corresponded to lower secondary education.

Regarding criminal history and sentencing, approximately half of the participants were serving their first custodial sentence, with an average time served ranging between one and four years. A total of 180 inmates (72%) had been convicted of physical violent offenses (including homicide, robbery, assault on public officials, and domestic violence), while 70 inmates (28%) had been convicted of sexual violent offenses (including rape, sexual assault, pimping, and consumption of child sexual abuse material).

The detailed distribution of participants according to offense type and sentencing characteristics is presented in

Table 1. Basic structure of the sample in terms of committed offenses and sentencing characteristics.

Variable		N	Percentage (%)
Type of violence	Sexual violence	70	28
	Physical violence—attempt/complicity	20	8
	Physical violence—robbery/assault on authority	120	48
	Physical violence—homicide	40	16
Number of convictions	First conviction	141	56
	Second conviction	41	16
	Multiple convictions	65	26
	NR	3	1.2
Sentence lengths (years)	1–4 years	107	42.8
	5–10 years	83	33.2
	11–15 years	19	7.6
	16–20 years	14	5.6
	21 years or more	24	9.6
	NR	3	1.2

Note: NR = non-response.

, which provides a clearer overview of the composition of the sample.

The distribution of inmates according to participation in social reintegration activities (

Table 2. Distribution of Inmates According to Participation in Social Reintegration Activities.

Activities	Participation
	Yes		No		Not Reported
	N	%	N	%	N	%
Schooling activities	92	36.8	140	56	18	7.2
Vocational training activities	81	32.4	149	59.6	20	8.0
Educational activities	203	81.2	33	13.2	14	5.6
Religious activities	193	77.2	47	18.8	10	4.0
Work–related activities	188	75.2	50	20.0	12	4.8
Psychological assistance activities	47 (daily)	18.0	197 (3–4 times/week)	78.8	8	3.2
Social assistance activities	43 (almost never)	17.2	200 (occasionally)	80.0	7	2.8

) indicated the highest levels of involvement in educational activities (81.2%). Fewer than half of the participants had engaged in formal schooling activities conducted within the penitentiary (36.8%), while approximately one third participated in vocational training programs (32.4%). With respect to psychological and social assistance activities, the majority of inmates reported occasional participation.

Table 3. Internal consistency analysis of the scales and subscales used in the present study.

Variable	No. of Items	Cronbach’s α	95% CI for Cronbach’s α
Overall aggression	29	0.894	0.874–0.912
Physical aggression	9	0.760	0.713–0.802
Verbal aggression	5	0.578	0.490–0.656
Anger	7	0.658	0.589–0.719
Hostility	8	0.811	0.774–0.845
Emotional stability	10	0.824	0.790–0.855
Dark Triad (Total)	27	0.818	0.783–0.849
Machiavellianism	9	0.802	0.763–0.837
Narcissism	9	0.443	0.333–0.541
Psychopathy	9	0.719	0.664–0.768

presents the internal consistency of the scales used in the study, indicating lower levels of internal consistency for the verbal aggression subscale and narcissism, and very good internal consistency for overall aggression and emotional stability.

Means, standard deviations, standard errors, minimum and maximum values, skewness, and kurtosis for each variable are presented in

Table 4. Descriptive analysis of the scales and subscales used in the present study.

Variable	N	M	SD	Md	Sk	K	KS	p
Overall aggression	250	66.57	18.90	65.00	0.63	0.26	0.06	0.030
Physical aggression	250	17.75	6.72	17.00	0.76	−0.01	0.11	<0.001
Verbal aggression	250	12.72	3.89	13.00	0.14	−0.41	0.08	0.001
Anger	250	14.79	5.04	14.00	0.65	0.23	0.12	<0.001
Hostility	250	21.30	7.21	21.00	0.34	−0.36	0.07	0.007
Emotional stability	250	34.52	7.19	34.5	−0.16	−0.19	0.05	0.200
Dark Triad (Total)	250	70.13	13.58	69.50	0.38	0.16	0.07	<0.009
Machiavellianism	250	24.27	6.90	25.00	−0.09	−0.26	0.09	<0.001
Narcissism	250	26.30	4.34	26.00	0.15	0.54	0.09	<0.001
Psychopathy	250	19.57	6.26	18.00	0.70	0.12	0.14	<0.001

Note: M = mean; SD = standard deviation; Md = median; Sk = skewness; K = kurtosis; KS = Kolmogorov–Smirnov test statistic for normality; p = significance level associated with the KS test.

. Overall, the results indicate that most participants exhibited moderate levels of emotional stability and aggression, with substantial interindividual variability in personality traits, including Machiavellianism, narcissism, and psychopathy. Reliability analyses indicated that most scales demonstrated acceptable (e.g., psychopathy) to very good (emotional stability) Cronbach’s alpha coefficients, with the exception of narcissism, which exhibited lower internal consistency.

Pearson correlation analyses (

Table 5. Results of the Pearson correlation analysis between Emotional Stability and Aggression.

Pearson Correlation (r)		Emotional Stability	Overall Aggression	Physical Aggression	Verbal Aggression	Anger	Hostility
Emotional stability	R	1	−0.481 **	−0.378 **	−0.291 **	−0.502 **	−0.417 **
	CI95% (LL, LU)	--	−0.586, −0.361	−0.496, −0.245	−0.420, −0.152	−0.604, −0.384	−0.530, −0.288
	N	180	180	180	180	180	180

Note: r = Pearson correlation coefficient; CI95% = 95% confidence interval for the correlation coefficient; LL = lower limit; LU = upper limit of the confidence interval; N = number of participants included in the analysis. ** p < 0.01 (two-tailed).

) indicated that emotional stability was significantly and negatively correlated with all dimensions of aggression: overall aggression (r = −0.48, p < 0.01), physical aggression (r = −0.38, p < 0.01), verbal aggression (r = −0.29, p < 0.01), anger (r = −0.50, p < 0.01), and hostility (r = −0.42, p < 0.01).

Correlation analyses were conducted on a subsample of 180 inmates convicted of physical violent offenses. These results support the first hypothesis, demonstrating that higher levels of emotional stability were significantly associated with lower levels of aggression across all its components, including physical aggression, verbal aggression, anger, and hostility.

The second hypothesis was tested using binary logistic regression analyses conducted to predict participation in social reintegration and work-related activities based on a model including the aggression subscales (

Table 6. Binary logistic regression analysis for predicting participation in schooling activities.

Step	−2 Log Likelihood		Cox & Snell R Square				Nagelkerke R Square
1	218.78 a		0.039				0.052
Variables in the equation	B	S.E.	Wald	Df	Sig.	Exp(B)	CI95%
Physical aggression	−0.08	0.04	4.59	1	0.032	0.93	0.86–0.99
Verbal aggression	0.04	0.06	0.36	1	0.548	1.04	0.92–1.17
Anger	0.04	0.05	0.66	1	0.417	1.04	0.94–1.15
Hostility	−0.02	0.03	0.51	1	0.474	0.98	0.92–1.04
Constant	0.49	0.62	0.62	1	0.432	1.63
Step	−2 Log likelihood		Cox & Snell R Square				Nagelkerke R Square
1	205.352 a		0.054				0.074
Variables in the equation	B	S.E.	Wald	Df	Sig.	Exp(B)	CI95%
Physical aggression	−0.09	0.04	5.84	1	0.016	0.91	0.85–0.98
Verbal aggression	0.05	0.06	0.39	1	0.535	1.04	0.92–1.18
Anger	0.09	0.05	1.99	1	0.159	1.08	0.97–1.20
Hostility	−0.06	0.045	1.84	1	0.175	0.95	0.89–1.02
Constant	0.50	0.64	0.60	1	0.437	1.65

Note: B = logistic regression coefficient; SE = standard error; Wald = Wald chi-square statistic; df = degrees of freedom; Sig. = significance level (p value); Exp(B) = odds ratio; CI95% = 95% confidence interval for the odds ratio; Cox & Snell R² and Nagelkerke R² = pseudo-coefficients of determination indicating the explanatory power of the logistic regression model; ^a = value for the final model including the constant.

). The results indicated that physical aggression was the only variable exerting a statistically significant effect on participation in schooling activities. Specifically, higher levels of physical aggression were associated with a lower likelihood of participation in these activities (OR = 0.93, 95% CI [0.86, 0.99]), indicating that each one-unit increase in physical aggression corresponded to approximately a 7% decrease in the odds of participation.

Physical aggression also negatively predicted participation in vocational training activities (OR = 0.91, 95% CI [0.85, 0.98]), corresponding to approximately a 9% decrease in the odds of participation for each one-unit increase in this dimension. The remaining components of aggression—verbal aggression, anger, and hostility—did not make statistically significant contributions to the predictive models.

The third hypothesis was tested using multiple linear regression analyses including the personality traits Machiavellianism, narcissism, and psychopathy as predictors. Separate regression models were estimated for each component of aggression (physical aggression, verbal aggression, anger, and hostility), as presented in

Table 7. Results of the linear regression analyses.

Aggression Components	Model Summary			ANOVA
	R	R2	Adjusted R2	F(3, 246)	p
Physical aggression	0.672	0.451	0.444	63.39	<0.001
Verbal aggression	0.410	0.168	0.158	16.55	<0.001
Anger	0.586	0.343	0.335	42.88	<0.001
Hostility	0.480	0.230	0.221	24.56	<0.001

.

All regression models were statistically significant (p < 0.001). The models explained 44.4% of the variance in physical aggression, 15.8% of the variance in verbal aggression, 33.5% of the variance in anger, and 22.1% of the variance in hostility.

According to the statistical results presented in

Table 8. Regression coefficients for each aggression component.

		B	SE	CI95% (B) (LL, LU)	β	T	p
Physical Aggression	Constant	4.75	2.01			2.36	−0.019
	Machiavellianism	0.10	0.05	−0.00, 0.21	−0.11	1.94	−0.053
	Narcissism	−0.11	0.08	−0.27, 0.05	−0.07	−1.40	−0.162
	Psychopathy	0.69	0.06	0.57, 0.80	−0.64	11.96	−0.000
Verbal Aggression	Constant	4.05	1.43			2.83	−0.005
	Machiavellianism	0.12	0.04	0.05, 0.20	−0.21	3.14	−0.002
	Narcissism	0.15	0.06	0.04, 0.26	−0.17	2.61	−0.010
	Psychopathy	0.09	0.04	0.01, 0.17	−0.15	2.27	−0.024
Anger	Constant	4.79	1.65			2.91	−0.004
	Machiavellianism	0.04	0.04	−0.05, 0.12	−0.05	0.86	−0.393
	Narcissism	0.01	0.07	−0.12, 0.14	−0.01	0.17	−0.862
	Psychopathy	0.45	0.05	0.36, 0.54	−0.56	9.53	−0.001
Hostility	Constant	2.45	2.55			0.96	−0.339
	Machiavellianism	0.32	0.07	0.19, 0.46	−0.31	4.73	−0.001
	Narcissism	0.35	0.10	0.15, 0.55	−0.21	3.38	−0.001
	Psychopathy	0.10	0.07	−0.05, 0.24	−0.09	1.35	−0.180

Note: B = unstandardized regression coefficient; SE = standard error; β = standardized regression coefficient; t = t-statistic; p = significance level; CI95% = 95% confidence interval for the unstandardized coefficient (B); LL = lower limit of the confidence interval; LU = upper limit of the confidence interval.

, psychopathy emerged as the strongest predictor of physical aggression, showing a significant positive association with physical aggression levels (t = 11.96, p < 0.001), whereas Machiavellianism exerted a marginal influence and narcissism did not demonstrate a statistically significant contribution. In the case of verbal aggression, Machiavellianism was identified as the strongest predictor (t = 3.14, p < 0.01), followed by narcissism and psychopathy. Regarding anger and hostility, psychopathy was a significant predictor of anger, while Machiavellianism and narcissism were significant predictors of hostility.

6. Discussion

The existing literature highlights the role of personality traits in adaptation to the prison environment, emphasizing in particular the importance of emotional stability as a protective factor against destructive and aggressive behaviors. Goodstein and Wright (1989) demonstrated that self-control and emotional stability facilitate institutional adjustment and reduce the risk of interpersonal conflicts. Subsequent research (Mehrabian 1997; Matthews et al. 2012; DeLisi et al. 2010) has further confirmed that emotional instability is associated with increased impulsivity and hostile reactions in stressful situations.

Within the carceral environment, this mechanism becomes more pronounced. Studies by Picken (2012) and Naidoo (2024) indicate that low levels of emotional control are associated with difficulties in complying with institutional norms and with an increased frequency of violent behaviors. More recent research (Grossi et al. 2023; Köhler et al. 2024) further confirms that psychopathy—particularly its affective component—is linked to reduced empathy and impaired emotional regulation, thereby accounting for the institutional adjustment difficulties also observed in the present study.

The findings of the present study are consistent with these observations, revealing significant associations between emotional instability and aggression, as well as between Dark Triad traits and difficulties in adapting to the carceral environment. Specifically, individuals exhibiting higher levels of psychopathy and Machiavellianism tended to display resistance to authority, behavioral rigidity, and lower engagement in social reintegration activities.

The first hypothesis was supported, demonstrating the existence of a significant negative relationship between aggression and emotional stability among inmates convicted of physical violent offenses. These findings are consistent with international literature highlighting emotional stability as a protective factor in the regulation of aggressive behavior. The meta-analysis conducted by Schenk and Fremouw (2012), which included more than 500 studies, confirmed that individuals exhibiting high emotional instability present a significantly increased risk of violent recidivism following release.

The findings of the present study further corroborate these observations, indicating that individuals with lower levels of emotional stability tend to respond impulsively to frustration and to perceive social interactions as threatening, which in turn facilitates hostile behaviors. This mechanism may be explained by deficits in emotional self-regulation characteristic of emotionally unstable profiles, whereby stress is managed through aggressive and poorly planned reactions.

It is important to note that aggression in the present study was assessed using the Buss–Perry Aggression Questionnaire (Buss and Perry 1992), which measures dispositional tendencies toward aggression rather than actual behavioral manifestations. Therefore, the results reflect trait aggression, understood as a relatively stable predisposition toward aggressive thoughts, emotions, and reactions, rather than directly observed acts of aggression. Distinguishing between dispositional aggression and observable aggressive behavior is important to avoid overinterpreting the findings as indicators of concrete aggressive actions within the institutional environment.

An important contribution of the present study is the validation of these relationships within the Romanian context, which remains underrepresented in the international literature, as well as in the identification of their practical implications. Early identification of emotional instability may guide psychological and psychoeducational interventions toward the development of self-control skills, frustration tolerance, and adaptation to institutional norms. Consequently, the relationship between emotional stability and aggression acquires not only theoretical relevance but also applied significance, with the potential to inform rehabilitation strategies within the penitentiary environment.

The second hypothesis, formulated to examine the extent to which aggression influences inmates’ participation in social reintegration activities, was only partially supported. The finding that aggression negatively predicts participation in social reintegration activities suggests that individuals with the highest violent potential are also the least receptive to correctional and social reintegration interventions. The analyzed data indicate that higher levels of aggression are associated with reduced involvement in educational and vocational training activities, whereas participation in other types of activities (moral-religious, cultural, or psychological counseling activities) does not appear to be significantly affected.

However, although physical aggression emerged as a statistically significant predictor, the magnitude of the observed effect was relatively small. Therefore, the practical implications of this finding should be interpreted with caution, as additional psychological and contextual factors are likely to influence inmates’ participation in reintegration activities.

This pattern may be explained by the fact that educational and vocational activities require rule compliance, cooperation, discipline, and self-control—behaviors that may be difficult to sustain for individuals exhibiting elevated levels of hostility and impulsivity. Inmates with pronounced aggressive traits tend to perceive such contexts as threatening to their group status, which may lead them to favor more individual-oriented activities. As highlighted by Goodstein and Wright (1989), motivation to participate in reintegration programs is contingent upon perceived utility and institutional trust factors that are often diminished among individuals with suspicious and hostile tendencies.

In addition, the findings concerning the role of the Dark Triad support the assumption that psychopathy is associated with direct aggressive behaviors, whereas Machiavellianism and narcissism are linked to more subtle forms of hostility and manipulation (Paulhus and Williams 2002; Muris et al. 2017). This differentiation among types of aggression underscores the complexity of inmates’ personality profiles and highlights the relevance of multidimensional assessment in tailoring intervention programs.

The present findings complement those reported by Chan and Beech (2024) and Lainidi et al. (2022), who identified associations between psychopathy and risk-taking behaviors, and between Machiavellianism and exploitative behaviors, respectively. These results further emphasize the instrumental nature of aggression within the carceral environment.

In this context, it is important to clarify that in the present study psychopathy was assessed as a personality trait within the Dark Triad framework using the Short Dark Triad Scale (SD3; Jones and Paulhus 2014). This operationalization differs from broader clinical or forensic conceptualizations of psychopathy assessed with instruments such as the Psychopathy Checklist (PCL) or the Levenson Self-Report Psychopathy Scale (LSRP). Although these measures share certain conceptual elements, they capture partially different aspects of the psychopathy construct. The Dark Triad measure reflects subclinical personality tendencies related to callousness, impulsivity, and manipulative interpersonal style, whereas instruments such as the PCL or LSRP are designed to assess a more complex and clinically oriented construct of psychopathy. Therefore, comparisons between the present findings and studies employing alternative psychopathy measures should be interpreted with caution.

The results obtained partially diverge from the conclusions of previous studies, which have reported a generalized refusal among aggressive inmates to participate in activities aimed at cognitive-behavioral change and social reintegration. Gendreau et al. (1999) explained this phenomenon by reference to the hostile prison environment, Crewe (2011) emphasized a lack of trust in authority, and Haney (2003) highlighted the desire to maintain an image of independence and dominance. From a similar perspective, Wooldredge and Steiner (2013) noted that participation in group-based activities requires a degree of openness and vulnerability that may be perceived negatively by aggressive inmates.

Consequently, these findings may be explained by the presence of selective motivation among inmates with aggressive profiles, who tend to choose activities perceived as relationally neutral while avoiding situations involving external control or emotional exposure. The results underscore the need to adapt educational and psychosocial programs to the psychological characteristics of these individuals by introducing gradual stages of engagement and by placing greater emphasis on the development of trust and self-control.

The third hypothesis was supported, indicating that psychopathy emerged as the strongest predictor of physical aggression, whereas Machiavellianism was the most salient predictor of verbal aggression. In addition, Machiavellianism and narcissism were identified as significant predictors of hostility. The relationship between Dark Triad traits and aggression has been consistently documented in studies conducted on incarcerated populations (Cale and Lilienfeld 2006), with findings similar to those of the present study showing that psychopathy is a significant predictor of physical aggression (Liu et al. 2021) and anger (DeLisi et al. 2010). Machiavellianism appears to play a particularly important role in predicting verbal aggression and hostility, as inmates may employ these forms of aggression instrumentally to consolidate their status within the prison hierarchy and to project authority and power in relation to others. Moreover, individuals exhibiting Machiavellian traits may be more prone to hostile and aggressive behaviors due to diminished feelings of guilt, regret, or remorse regarding the consequences of their actions. In this sense, hostility—consistent with the findings of the present study—may represent an adaptive and instrumental strategy for coping with the penitentiary environment.

The findings of the present study carry important practical implications for the field of correctional psychology. A comprehensive understanding of inmates’ personality profiles may guide rehabilitation programs toward more individualized interventions focused on enhancing emotional self-regulation and reducing impulsive behaviors. In particular, the identification of dysfunctional Dark Triad traits may contribute to the development of differentiated strategies for counseling, training, and post-release integration. As suggested by Grossi et al. (2023), interventions aimed at increasing prosocial behavior and reducing impulsivity have the potential to mitigate the impact of psychopathic traits.

Accordingly, rehabilitation programs designed for inmates exhibiting high levels of psychopathy should incorporate components targeting emotional regulation, empathy, and the development of interpersonal skills. Attention should be given to individuals with severe psychopathic traits, as these may reduce the effectiveness of traditional intervention programs. Regarding empathy, a focus on cognitive empathy—namely, understanding the consequences of aggressive and criminal behaviors—may be more feasible, given that profound emotional change is often difficult to achieve in this population.

From an institutional perspective, these findings support the inclusion of periodic psychological assessments within intervention plans to monitor the progression of emotional and behavioral adjustment throughout the period of incarceration.

The findings of the present study may also be interpreted within a broader conceptual framework linking personality traits, emotional regulation, and institutional adjustment. Emotional stability reflects individuals’ capacity to manage stress and regulate emotional responses, which may contribute to lower levels of aggression within the institutional environment. In turn, reduced aggression may facilitate greater engagement in structured activities that support social reintegration. At the same time, Dark Triad traits were associated with higher levels of aggression, suggesting that certain maladaptive personality characteristics may indirectly influence patterns of institutional engagement through their relationship with aggressive behavior. From this perspective, aggression may function as an intermediate behavioral mechanism linking personality characteristics with patterns of participation in reintegration activities.

Limitations and Future Research Directions

Several limitations of the present study should be acknowledged. One important limitation concerns the provision of institutional credits to inmates for completing the questionnaires. Within the Romanian penitentiary system, such credits are granted for participation in educational, rehabilitative, or research-related activities and may contribute to recommendations for certain rewards in accordance with legal regulations. Although these credits do not represent direct monetary compensation, their provision may have influenced inmates’ willingness to participate in the study and could raise questions regarding participants’ motivations for consenting to take part in the research.

Another limitation of the present study concerns the heterogeneity of the sample in terms of offense type. Violent offenders do not represent a homogeneous group with respect to personality traits, motivations, or behavioral patterns. Previous research has highlighted substantial variability across different subtypes of violent offenders (Herrero et al. 2018; Shimotsukasa et al. 2019; Chow et al. 2025).

In the present study, participants convicted of different forms of violent offenses—including homicide, robbery, and sexual violence—were analyzed within broader categories due to the practical difficulty of accessing large and homogeneous samples in correctional settings. This heterogeneity may limit the possibility of drawing conclusions specific to subgroups of violent offenders. Future research should therefore examine these subtypes separately to better understand the role of personality traits in different forms of violent behavior.

The partial confirmation of one of the hypotheses indicates the need for future research to expand the sample to a larger number of participants drawn from multiple penitentiary institutions, encompassing different custodial regimes and varying levels of involvement in productive and social reintegration activities.

The present study also highlights the limitations associated with administering questionnaires to inmates by a representative of institutional authority, as well as the need to consider the inclusion of interviews with professionals directly involved in delivering specialized interventions to the study participants. Future research could integrate qualitative methods—such as semi-structured interviews, behavioral observations, or assessments provided by penitentiary staff—in order to capture more accurately the dynamics of adaptation and interpersonal relationships within the carceral environment. Moreover, extending the investigation at a national level through diverse samples would allow for comparisons across detention regimes and facilitate the validation of predictive models of penitentiary adjustment.

Despite these limitations, the present study provides a meaningful contribution to the understanding of the psychological mechanisms involved in adaptation to the penitentiary environment. The findings may serve as a starting point for the development of differentiated psychological assessment and intervention programs focused on personality traits and aggression management, thereby contributing to the reduction in recidivism risk and to the enhancement of social reintegration processes.

7. Conclusions

The results of the present study revealed a significant negative relationship between emotional stability and aggression, indicating that inmates with lower levels of emotional control are more likely to exhibit impulsive and hostile reactions. This finding has important practical implications for psychological assessment and for the selection of inmates for social reintegration programs, underscoring the need to include measures of emotional stability within penitentiary clinical screening procedures.

High levels of aggression were associated with reduced participation in educational and vocational activities, but not in moral-religious or counseling activities. The finding that aggression decreases the likelihood of engagement in reintegration programs—particularly those delivered in group settings—highlights a major obstacle in the rehabilitation of violent offenders and represents a key contribution of the present study. These results emphasize the necessity of adapting specialized intervention programs to the psychological profiles of aggressive inmates.

Psychopathy was primarily associated with direct forms of aggression, whereas Machiavellianism and narcissism were linked to more subtle, manipulative, and hostile manifestations. Inmates exhibiting these traits demonstrated greater difficulties in adaptation and interpersonal functioning, confirming the role of Dark Triad dimensions as behavioral risk factors.

Therefore, personalized psychotherapeutic interventions should focus on impulse control, the development of empathy, and the enhancement of socio-emotional competencies in order to facilitate adaptive functioning within the penitentiary environment and, subsequently, within the community.