Introduction

The most recent editions of major diagnostic systems, including the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition [1] and the International Classification of Disease, 11th Revision [2], have included the specifier with Limited Prosocial Emotions to describe a subgroup of young people with severe behavioral issues who have high levels of CU traits. CU characteristics include a shallow or limited affect, lack of empathy, lack of regret, and a lack of concern for one’s success in important tasks, according to Frick [3].

According to theory, CU traits are an affective aspect in children [4] and the larger construct of psychopathy in adults [4]. Additionally, studies showing links between CU characteristics and especially serious (i.e., aggressive) and persistent types of antisocial behaviour supported their inclusion in the clinical classification [3].

It has been demonstrated, for instance, that young people with higher CU features use weapons more frequently [4,5] and exhibit more destructive types of violence [6–9]. Furthermore, even when there are no significant conduct issues, youth with higher CU features are observed to have more psychosocial and relationship dysfunction [10–12]. For instance, it has been demonstrated that young people with higher CU traits face more rejection from their peers and have a greater likelihood of being labelled as “mean” by them [13–15]. Moreover, CU characteristics appear to indicate an etiologically unique group of young people with serious behavioural issues who exhibit less sensitivity to punishment and less emotional responsiveness to the pain of others [16].

According to [17], however, it seems that roughly 25% to 30% of young adolescents with serious conduct issues had elevated levels of CU characteristics. Thus, it’s critical to evaluate these characteristics in a valid and reliable way. Despite their size, this subset of children and teens needs to be evaluated and treated professionally for a number of crucial reasons, such as the fact that their behavioural issues are more serious and persistent. In this instance, they also exhibit an increased level of violence, which is more goal-oriented—for dominating the other individual or for personal gain—and tends to injure the victims more [18,19]

The Psychopathy Check List-Youth Version (PCL-YV) [20] and the Antisocial Process Screening Device (APSD) [21] are two frequently used tools in research due to the significance of evaluating individuals for their CU features, especially with conduct problems. The PCL-YV’s drawback is the amount of time required to complete the assessment; even highly qualified professionals need about 60–90 minutes. As for the APSD, it is difficult to verify the significant aspects of CU characteristics because only six of the 20 items on the APSD assess CU characteristics, which can be self-reported by young people, their guardians, or their instructors [21]. The inventory of CU traits was developed by Frick [22] to expand the CU characteristics items on the APSD in response to these shortcomings of other assessments. It has 24 items with 4-point Likert scale codes. Versions of the Youth Self-Report, Parent Report, Teacher Report, and Parent Report (Preschool) are available. Callousness, which describes actions that include a lack of empathy, guilt, and regret; Uncaring, which denotes a lack of concern for one’s performance on tasks and for the feelings of others; and Unemotional, which denotes a lack of emotional expression, are the three main characteristics of CU that have been found in studies [23].

Various samples with varying translations have usually supported these three characteristics [24–26]. For instance, these three factors were discovered for the first time by Essau et al. [23] using an exploratory factor analysis in a sample of 1443 Germans aged 13–18 from the general population. They also discovered a common general characteristic that included all of the items (i.e., the bifactor model) of the self-report version, and the factor structure was the same for both boys and girls. Using the self-report version, Kimonis and colleagues [22] validated this factor structure in a sample of 248 American juvenile offenders aged 12–20. The Unemotional factor’s internal consistency was low (0.53), and two Callousness items (items 2 and 10) were eliminated from their model because of poor item-total correlations. The bifactor model was validated using a self-report version in 540 Italian children (mean age = 12.7 years) by Ciucci and colleagues [12], and the factor structure was consistent across grade and sex.

A few recent studies have questioned the validity of scales and the bifactor model [27–29]. Willoughby and colleagues [29] examined the parent ratings of the ICU in 1078 American first-graders (mean age = 7.3 years) and discovered that the best fit was a two-factor model that distinguished callous-unemotional (including nine Callousness and two Unemotional items from the original scale) and empathic-prosocial (including eight Uncaring, two Callousness, and three Unemotional items) factors. Furthermore, not a single item was cross-loaded on more than one variable.

In a study of 250 American teenagers with significant conduct problems (ages 6–12), Hawes and colleagues [28] developed a structure based on two factors (callousness and uncaring) using 12 of the original 24 items of the parent-report version. In a research by Houghton and colleagues [30] that examined the factor composition of self-report ICU in 268 Australian children aged 7–12, a two-factor model worked better than a three-factor model. The two-factor model’s unemotional items were all removed due to low internal reliability and model fit. Each of the eight items in the Callousness and Uncaring categories exhibited internal reliabilities of 0.77 and 0.85, respectively, suggesting sufficient reliability. Eight pairs of error terms were connected in order to improve model fit even more.

Using 613 Chinese males who were detained among the Chinese samples, Zhang and colleagues [31] investigated the factor structure and psychometric properties of the Chinese equivalent of the ICU. According to the findings of the confirmatory factor analysis, the three-factor model that included 24 items from the original ICU scale did not show a good match to the data. However, the two-factor with factors of callousness and uncaring—which included only eleven items—showed the most correlation with the data. The Youth Psychopathic Traits Inventory and the Antisocial Process Screening Device were also compared with the two-factor ICU in the study, and both instruments showed comparable and noteworthy correlations with the relevant criteria of violence, psychopathy, and compassion.

In a different study, Allen and colleagues [32] evaluated the invariance between the Chinese and British populations using both the original 24 items and the shorter ICU (11 items). The study comprised 364 Chinese children, aged 10–13, who were studying in schools in Guangdong, China, and 437 British students, aged 11–14, from the East of England. Both the original 24-item and the 11-item ICUs showed strong internal consistency for the sub-factor and overall scores, but not for the unemotional factor in either sample. The unemotional measure has demonstrated low internal consistency and poor construct validity in previous research, per a recent meta-analysis study [33].

Although the ICU has been widely validated across different cultural contexts, findings regarding its factor structure remain inconsistent. While early studies generally supported the original three-factor or bifactor structure, more recent research has increasingly questioned the reliability and distinctiveness of the Unemotional dimension. Several cross-cultural studies have instead reported improved fit for shortened two-factor models consisting primarily of Callousness and Uncaring dimensions. These inconsistencies suggest that the underlying structure of CU traits may vary across linguistic and cultural contexts, as cultural norms surrounding emotional expression and interpersonal behavior may influence how CU traits are interpreted and reported across populations [34,35], highlighting the need for further validation studies in underrepresented populations such as Urdu-speaking youth.

Therefore, the present study contributes not only to the linguistic adaptation of the ICU into Urdu but also to the broader theoretical discussion regarding the cross-cultural stability of CU trait structures. By comparing competing factor models within a Pakistani sample, the study aims to clarify whether previously reported two-factor solutions generalize to Urdu-speaking populations. Despite the widespread international adaptation and validation of the ICU, no validated Urdu version currently exists. The precise diagnosis and treatment of conduct disorder and psychopathic features in Urdu-speaking populations, especially among Pakistani adolescents and young adults, is severely restricted by this shortcoming. Therefore, the present study aimed to translate, culturally adapt, and examine the factor structure and psychometric properties of the Urdu version of the Inventory of Callous-Unemotional Traits (ICU-U) in a Pakistani sample.

The construct validity of the ICU has been supported by numerous studies that show a strong correlation between CU traits and violent, delinquent, and antisocial conduct [18,22,36]. Nevertheless, no research has been done to examine the convergent and discriminant validity of the ICU in Urdu. By testing the Urdu version’s convergent and discriminant validity, we address this gap.

The primary objective of the research was to examine the psychometric qualities of the self-report ICU in young people selected from the community. In particular, (1) we will use a series of CFAs to compare competing measurement models previously examined in the literature, including an intercorrelated three-factor model, a three-factor bifactor model, and a two-factor model, and (2) we will examine the scale’s convergent and discriminant validity. We anticipated that this sample would duplicate the two-factor structure (Callousness and Uncaring) based on previous research [37].

Method

There were three phases in this study. A multi-phase design was adopted to ensure comprehensive cross-cultural validation of the ICU-U. The first phase focused on linguistic and cultural adaptation of the instrument, the second phase examined temporal stability across language conditions, and the third phase evaluated the structural and construct validity of the translated scale. This sequential approach is consistent with recommendations for psychometric validation studies and allowed the study to examine multiple aspects of reliability and validity within the Urdu-speaking context. The details of the three phases are below.

Results

The results of all phases are described below.

Results of test-retest reliability

The test-retest reliabilities of the translated (Urdu) version of the study instrument were established using the data gathered from young people using the previously described approach. The following are the results that were obtained.

All four groups have positive test-retest reliability and statistical significance, as shown in Table 1. In accordance with the criteria for test-retest reliability set by Sparrow and colleagues [45], good to excellent test-retest reliability is indicated by Pearson correlations and intraclass correlation coefficients (ICC; absolute agreement) above  .53 (<.40 = poor,  .40 –  .59 = fair,  .60 –  .74 = good, and  .75–1.00 = excellent). However, given the small sample size in each retest group (n = 15), these findings should be interpreted as preliminary evidence of temporal stability. Since ratings remain constant over time, these findings confirmed that they possessed external consistency and temporal validity within the sample. The results also showed that, of the four groups, the Urdu–Urdu group’s correlation coefficients for the scale were stronger and higher than the English–English group’s. This suggests that, in comparison to the English version, the translated Urdu form was found to have greater understanding. These results demonstrated that young male Pakistani people were able to validate the translated version of the ICU across languages.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Test–Retest Reliabilities of English and Urdu Version of the Inventory of Callous Unemotional Traits (N = 60).

https://doi.org/10.1371/journal.pone.0353300.t001

Results of psychometric properties

Confirmatory factor analysis.

Table 2 shows the fit indices of competitive models used in the current study. Fit indices showed an unacceptable fit for the inter-correlated three-factor model (M1; χ² = 1255.82, df = 249, CFI = 0.77, TLI = 0.74, RMSEA = 0.12) and for the original three-factor bifactor (M2; χ² = 1255.81, df = 249, CFI = 0.76, TLI = 0.74, RMSEA = 0.11). The two-factor model of the ICU 12 had significantly better fit than the M1 or M2, but the fit indices were still unsatisfactory (M3; CFI = 0.89, TLI = 0.86, RMSEA = 0.08). After adding error variance, the values of the GFI, IFI, CFI, χ²(df), and RMSEA fell into the acceptable range. A good fit was indicated in Model 3 (adjusted) of ICU 12.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Goodness-of-fit Indices for the Different Models of ICU-U (N = 300).

https://doi.org/10.1371/journal.pone.0353300.t002

In a two-factor model, the factor loading for the ICU-U was displayed in Table 3. Every item displayed factor loading values above the cutoff point (.30).  .41 to  .73 is the range of factor loadings for the two-factor model.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Factor Loadings for the Two-factor Model for ICU-U 12 (N = 300).

https://doi.org/10.1371/journal.pone.0353300.t003

Convergent and discriminant validity.

Convergent and Discriminant validity were conducted using JASP version 0.19. The analysis of the results revealed ICU-U was positively (r = .84, p < .01) related to CSS-M and negatively (r = −.84, p < .01) with the Brief Self-Control Scale. Results indicate that CU traits are related to greater criminal attitudes, which attests to convergent validity. Similarly, CU traits and self-control are strongly and negatively correlated, which indicates that greater CU traits are related to less self-control, which attests to discriminant validity.

Discussion

The current study aimed to translate, culturally modify, and evaluate the ICU in Urdu among young people in Pakistan. The findings provide the ICU-U with excellent psychometric support, demonstrating its validity and reliability as a tool for evaluating callous-unemotional (CU) qualities in Urdu-speaking cultures. Given the lack of validated instruments in Urdu for identifying young people with Limited Prosocial Emotions and associated behavioral issues, this is a noteworthy contribution.

In Phase I, we translated ICU using the forward-backward translation approach in accordance with Brislin’s [38] recommendations. No ambiguities, contradictions, or conceptual inconsistencies were identified between the original and back-translated versions. As a result, the Urdu version of the ICU was judged to demonstrate adequate linguistic and conceptual equivalence and was considered suitable.

Test-retest results were favorable. All test-retest groups showed strong reliability (ICC values between  .53 and  .84), with the Urdu–Urdu condition showing the strongest reliability (r = .85; ICC = .83). This shows not only strong temporal stability but also that the Urdu version may provide greater clarity and cultural resonance for native speakers compared to the original English version. These results are consistent with cross-national research highlighting the importance of culturally appropriate instruments [38,40].

Results of the CFA indicated that the first proposed inter-correlated three-factor and bi-factor models did not fit the data in the Pakistani sample [3,22]. Instead, a more nuanced two-factor model including Callousness and Uncaring fitted better, consistent with recent findings in several cultural contexts [28–30]. This finding may indicate that the affective and interpersonal dimensions of CU traits are more readily identifiable and culturally interpretable within collectivistic contexts such as Pakistan, whereas unemotionality may be expressed differently across cultures. These findings support growing evidence that the structure of CU traits may not be universally stable across cultural contexts and highlight the importance of culturally sensitive psychometric validation.

Our model had 12 items, with all factor loadings above the minimum threshold of 0.30, while these only had 11 items and a two-factor model [28–30]. We wanted to ensure that the factor structure of the ICU was adequate for the local population [31]. There are occasions when specific subscales and items have either been changed or removed in studies as a function of their psychometric performance; this includes modifications so that the scale is more appropriate for the underlying population. This methodological decision is consistent with previous cross-cultural psychometric validation studies which have deleted some items or subscales based on a lack of relevance or lower psychometric properties.

The internal consistency of the two factors was satisfactory (α = .76 for Callousness and α = .82 for Uncaring), showing that the ICU-U reliably measures these two factors of CU. The poor fit and low reliability of the Unemotional sub-scale, consistent with prior research [32–33], led to its removal from the final model. This supports the recent consensus that unemotionality may not be psychometrically distinct or consistently measurable across cultures and age groups.

The ICU-U also demonstrated excellent convergent validity by showing a strong positive relation with the Criminal Sentiments Scale – Modified (r = .84, p < .01), confirming that CU characteristics are related to pro-criminal attitudes. This supports previous results that CU characteristics predict antisocial and aggressive behaviors [18,19,22]. From a theoretical perspective, these findings align with developmental psychopathology models suggesting that CU traits are associated with reduced emotional responsiveness, impaired empathy, and diminished behavioral regulation, all of which contribute to antisocial attitudes and self-regulatory difficulties [19,22]. The strong associations observed in the present study further support the construct validity of the ICU-U within the Pakistani cultural context.

Moreover, the ICU-U showed strong discriminant validity, as shown by a significant negative correlation with the Brief Self-Control Scale (r = –.84, p < .01). This result is theoretically sound, as lower self-control is a known correlate of CU characteristics and psychopathy [43,44]. These findings enhance confidence in the scale’s ability to differentiate CU traits from other, distinct psychological constructs.

Beyond providing a translated instrument, the present study contributes to the broader literature on CU traits by demonstrating that psychometric structures established primarily in Western populations may not fully generalize across cultural and linguistic settings. The findings, therefore, emphasize the importance of examining the cultural variability of CU trait expression and measurement.

Conclusion

In summary, the Urdu version of the ICU-U demonstrates good psychometric properties, including preliminary evidence of temporal stability, a robust two-factor structure, high internal consistency, and strong convergent and discriminant validity. This makes the ICU-U a culturally sensitive and reliable questionnaire for evaluating CU traits in the Urdu-speaking sample, with wide implications for research, diagnosis, and intervention planning.

Supporting information

Acknowledgments

The authors thank all participants for their time and contribution to this study. We also acknowledge the assistance of colleagues who supported the translation and data collection processes. The authors would also like to express their sincere gratitude to the Tohoku University Support Program for the Article Processing Charge (APC) FY2026 for their generous financial support, which made the publication of this work possible. Their commitment to advancing academic research and promoting the dissemination of knowledge is greatly appreciated.