Emotion Recognition and Perceived Social Support in Young People Who Offend

Abstract

Young people who offend appear stuck in a cycle of adverse experiences, low levels of social support and emotional skill deficits. Yet these factors have not been extensively researched with young people who offend. The current study aimed to develop the understanding of emotion recognition ability and perceived social support in young people who offend and to explore the relationship between these variables. A total of 50 young people who offend were recruited through three Youth Offending Teams and 50 age, gender, ethnicity, socio-economic status and academically matched young people without a known offending history were recruited from a college and youth service in the same geographical area. All participants completed a demographic questionnaire, the Toronto Alexithymia Scale, a Facial Emotion Recognition Task, a Verbal Emotional Prosody Recognition Task and the Multidimensional Scale of Perceived Social Support. Failing to support the hypotheses, statistical analyses failed to show that, relative to the controls, young people who offend had significantly higher levels of alexithymia, lower levels of perceived social support or lower ability to recognise others’ emotions. However, relative to the controls, young people who offend did show significantly lower ability to recognise fear through verbal prosody. Of particular interest, looked after status, which was more commonly reported among young people who offend (38%) than controls (4%), was the predominant factor associated with all outcome variables. Thus looked after status, rather than offending status in isolation, is more associated with difficulties in identifying and describing feelings, ability to recognise others’ emotions and levels of perceived social support. In addition, significant correlations were found between (1) alexithymia and perceived social support, (2) the ability to recognise others’ emotions and perceived social support and (3) the ability to recognise emotions from facial expressions and verbal prosody. Theoretical and clinical implications of the study findings are discussed and areas for future research are suggested.

Keywords

alexithymia emotion recognition facial emotion recognition looked after children perceived social support verbal emotional prosody recognition young people who offend

Introduction

Background

Criminal justice systems around the world are based heavily on the use of deterrence, restorative justice and rehabilitation (Andrews and James, 2010; Bergseth and Bouffard, 2013). However, young people who offend indicate that their needs are being ignored and poorly met by professionals (Chitsabesan et al., 2006; UserVoice, 2011) and many interventions for young people who offend are focused on anger management and victim empathy, which assume that young people who offend are able to draw on socially supportive relationship skills and recognise and express emotions. However, a review of the literature suggested that young people who offend are stuck in a cycle of adverse experiences and deficits in emotional skills and socially supportive relationships.

Emotional skills and social support

Emotional skills are described as ‘the ability to monitor one’s own and others’ emotions, to discriminate among them and to use the information to guide one’s thinking and actions’ (Salovey and Mayer, 1990: 198). Research consistently reports a relationship between emotion recognition from faces and emotion recognition through tone and intonation (also known as prosody) verbal prosody (Mill et al., 2009; Rigoulot and Pell, 2014). Studies also report a relationship between recognising one’s own emotions and recognising others’ emotions (Israelashvili et al., 2019), with deficits in recognising and labelling one’s own emotions (alexithymia), related to difficulty in associating emotional cues from facial expressions or verbal emotional prosody to others’ emotional states (Grynberg et al., 2012). Emotional skills are also thought to be essential to the development of friendships and social support (Fitness, 2006).

The majority of social support tends to be informal, arising through friends and family (Leach, 2015). Social support can provide varied types of resources (emotional, practical or informational), aiding the recipient’s health and well-being and their ability to cope with stressful life events (Lakey and Cohen, 2000), although emotional skills are thought to mediate the impact of social support on well-being (Chen and Feeley, 2012). Social support can be divided into three elements: (1) social connectedness/embeddedness; (2) perceived social support and (3) actual or enacted social support (Barrera, 1986; Burleson and MacGeorge, 2002; López and Cooper, 2011; Sarason et al., 1990). Studies report a significant relationship between alexithymia and perceived social support (Karukivi et al., 2014) and facial emotion recognition (FER) and perceived social support (Collin et al., 2013). And, difficulty in relationships can result in problems in reliably expressing and labelling emotions (Spitzer et al., 2005).

It is thought that cognitive processes involved in emotional skills, social behaviour and the effective management of situations and relationships are related to early attachment relationships and experiences (Fox et al., 2010). Impaired or underdeveloped emotion recognition skills are associated with early attachment difficulties when children’s internal experiences are not understood and labelled by others (Meins et al., 2002; Murray and Andrews, 2005). Early (attachment) experiences have also been thought to affect everyday appraisal, memory of and attention to support (Integrated Model of Perceived Social Support (IMPSS), Sarason et al., 1990). Furthermore, different attachment styles are related to the availability of socially supportive relationships, social support skills (Ma, 2006), willingness to seek social support (Bartholomew et al., 1997) and feelings experienced at receiving support (Sarason et al., 1990).

The majority of young people who offend are exposed to early adversity (also known as adverse childhood experiences) (Chitsabesan et al., 2006) and nearly half of young people who offend in institutions have spent time in local authority care (Blades et al., 2011). For such children, early secure attachments and positive experiences are less likely than children who have not been in local authority care, endangering healthy neurological development (Biehal et al., 2010; Fox et al., 2010). Indeed, children leaving local authority care show higher levels of emotion recognition deficits (Barone and Lionetti, 2012; Hollingworth, 2014) and have a lower availability of socially supportive relationships (Ma, 2006) and positive feelings experienced at receiving support (Sarason et al., 1990).

Difficulty understanding one’s own or others’ emotions makes communication problematic and can increase the risk of anti-social and impulsive behaviours associated with offending (Fossati et al., 2009). Theoretically, it is proposed that a reduced ability to recognise negative emotions (particularly fear) can lead to reduced feelings of punishment by others’ fearful expressions, reducing the likeliness of an empathic response and inhibition of the behaviour that caused this distress (Bird and Viding, 2014; Blair, 2005; Murray and Andrews, 2005). The majority of emotion recognition studies with young people (some of who have offended) have explored FER difficulties in relation to psychopathy (Blair et al., 2001; Dadds et al., 2006), callous unemotional traits (Jusyte et al., 2014; Wolf and Centifanti, 2014), conduct disorder (Fairchild et al., 2009) and anti-social behaviours (Marsh and Blair, 2008). A review of the literature completed for the purposes of this study identified a number of studies with a specific focus on FER deficits and offending behaviour among young people. These studies report both overall FER deficits (Gonzalez-Gadea et al., 2014; Schofield et al., 2015) and specific FER deficits for negative emotions such as sadness (Bowen et al., 2013; McCown et al., 1986), anger (Bowen et al., 2013; Jones et al., 2007), disgust (Jones et al., 2007; McCown et al., 1986; Sato et al., 2009) and fear (Bowen et al., 2013). Subgroup analysis of young people who offend samples revealed a significant negative correlation between FER and violent offences (Carr and Lutjemeier, 2005).

A number of studies were also identified exploring recognition of emotions through verbal emotional prosody and offending, although these included exploration of emotion recognition deficits in relation to psychopathic traits (Dawel et al., 2012) in adult forensic samples (Bagley et al., 2009; Suchy et al., 2009) and pupils attending schools which support social, emotional and behavioural needs (Blair et al., 2005). These studies all report a specific deficit in recognising fear from vocal cues among these participants. Despite the lack of research with young people who offend, one might hypothesise that young people who offend will show similar deficits, based on findings with individuals presenting with similar difficulties (Dawel et al., 2012), reports of significant correlation between emotion recognition through facial expressions and verbal prosody (Rigoulot and Pell, 2014) and young people who offend showing significant FER deficits.

Alexithymia (literally meaning ‘no words for feelings’, Taylor et al., 1997) has also been reported to be significantly correlated with offending behaviour among adolescents (Berastegui et al., 2012) and with psychopathic traits in young people who offend (Langevin and Hare, 2001). Although some studies have reported young people who offend to show higher alexithymia than the controls (Möller et al., 2014; Moriarty et al., 2001; Zimmermann, 2006), these differences only reached statistical significance in one study (Zimmermann, 2006). Subgroup analysis revealed non-significant higher levels of alexithymia among violent offenders than non-violent offenders (Möller et al., 2014), with alexithymia related to a reduced ability to regulate one’s emotions, increasing the risk of violent expression of emotional states (Fossati et al., 2009).

Research examining the effect of social support on offending behaviour is also limited. There are reports that while young people who offend are detained, social support from nonparent family members (Johnson et al., 2011), visits from parents (Monahan et al., 2011) and emotional and practical peer support (Bagnall et al., 2015) significantly reduce the experience of psychological distress. Family ties are reported to increase emotional well-being and reduce housing and employment issues (Ministry of Justice [MoJ], 2008; Ministry of Justice and Department for Children, Schools and Families, 2009), and are pivotal to successful community reintegration (James et al., 2013). However, young people who offend residing in the community have significantly higher unmet social support needs than those residing in secure settings (Chitsabesan et al., 2006). Young people who offend residing in the community are most likely to seek and receive support from people with whom they have a long-standing relationship. Of note, some socially supportive friendships can be erratic and unpredictable in nature (Colvin et al., 2002) and criminally embedded (Clear et al., 2001). It is hardly surprising, then, that social support can also have negative outcomes for young people who offend, with context, source and dimensions of support, especially qualities possessed by those providing support (e.g. unrealistic family expectations and peer temptation), affecting outcome (Leach, 2015; Martinez and Abrams, 2013; Salvatore and Markowitz, 2014). Young Offending Services are a less preferred source of support, not as a result of a lack of provision, but because of psychological, social, structural and cultural barriers (King et al., 2014; Walsh et al., 2011).

Limitations of the literature and study aims

Despite the literature identifying reduced social support and emotional skills as risk factors for offending behaviour, these factors remain relatively unstudied with adolescents and do not appear to have been studied with a sample of young people who offend. Only one FER study has been reported with a community sample of young people who offend (Bowen et al., 2013). Alexithymia is less commonly studied with young people who offend and no studies were identified exploring emotion recognition from verbal prosody with young people who offend. Only three of the emotion recognition studies with young people who offend considered the impact of offending variables (offence type and severity) and all studies were completed with male participants. No studies to date have measured both alexithymia and the ability to recognise others’ emotions in a sample of young people who offend, and no studies have measured perceived social support in a sample of young people who offend. A number of limitations also need to be considered when interpreting research findings, such as lack of sample matching and controlling for potentially confounding variables.

In order to develop effective policy and practice with respect to reducing the risk of offending and re-offending among young people, it is vitally important to better understand the predictors and complex interplay of the different needs of young people who offend. The current study aimed to build on the research findings to date, by being the first study to measure both emotion recognition and perceived social support with young people who offend, including males and females, while controlling for any confounding variables and completing within group analyses. It was hypothesised that, relative to controls, young people who offend would show significantly lower levels of emotion recognition and lower levels of perceived social support. Further aims of the study were to explore the relationships between the various outcome variables.

Methodology

Participants

The study group (N = 50) consisted of males (N = 38) and females (N = 12) aged 14–18 years (mean = 16.32 years), in current contact with any of the three South-East Wales Youth Offending Teams and with capacity to provide informed consent themselves (and via parent/guardian if aged 14–15). Young people in the study group had committed a range of different offences at different levels of frequency and severity. The recruitment process involved case workers discussing the research with young people who were on their caseload. If young people expressed an interest, case workers provided the researcher with the young persons’ contact details (with young people’s consent).

The controls (N = 50) consisted of males (N = 38) and females (N = 12) aged 14–18 years (mean = 16.24 years) and had capacity to provide informed consent (as well as parent/guardian consent if aged 14–15). The control group was recruited by discussing and agreeing project details with relevant personnel from local education and youth services. Potential participants were provided with research information and inclusion/exclusion criteria through staff communication and study posters. If young people expressed an interest, staff informed researchers of the most suitable times and locations for the young people to participate. Every effort was made for ccontrols to be allocated appropriately to this group (hadn’t committed any known offences). Staff and participants were made aware of the exclusion criteria (i.e. known to have engaged in offending behaviour) in the recruitment process and participants were asked whether they had ever committed an offence or been in trouble with the police during a pre-participation interview.

Service user involvement and well-being

All aspects of the research reported here were scrutinised and approved by the Cardiff University School of Psychology Research Ethics Committee. A group of young people (with and without a known offending history) were consulted prior to commencement of the study in relation to the study assessment measures, materials for recruitment and debrief, questionnaire format and administration. This information was used to modify the design of the research methodology and the choice of materials. For example, a clear explanation was provided around privacy of data, reasons for participation and the personal nature of some questions, which could also be left blank by selecting ‘prefer not to say’. The questionnaire and all written information were enhanced with colour and pictures. To reduce difficulties related to dyslexia and other reading deficits, participants were given the option to wear headphones throughout questionnaire completion to listen to recorded verbal instructions alongside the written instructions.

Although participation in this study was not anticipated to cause distress, ethical considerations were given careful attention, given participants age and potential vulnerability. Participants were invited to ask questions throughout their participation if they did not understand anything or required further support. In addition, a period of time was allocated to debrief participants after completion of the measures to discuss any concerns that may have arisen and an accessible debrief form, containing support service information, was also given to all participants.

Materials

All participants completed all five self-report measures through the Medialab software package (Empirisoft Corporation, New York). Measures included a demographic questionnaire, a perceived social support measure, an alexithymia measure and two measures of emotion recognition of others, including an FER task and verbal emotional prosody recognition (VEPR) task. For the young people who offend group, the demographic questionnaire also included a section for participants to indicate what offence(s) they had committed.

Demographic questionnaire

The demographic questionnaire asked participants to report on age, gender, ethnicity, academic achievement (grades on academic work and qualifications), whether they had ever spoken to a professional about emotional difficulties and whether they had ever spent time in care. Socioeconomic status (SES) was estimated using the UK’s Office for National Statistics estimates of average household weekly income based on the participant’s postcode (low = £0–£520; middle = £521–£670; high = £671+).

The young people who offend group also completed a multiple-choice self-report measure of offences committed, informed by the Youth Justice Board Counting Rules March 2006–April 2007 (as cited by Bowen et al., 2013). Participants were given the choice to ‘prefer not to say’ and also to select and specify ‘other’ offences, if their offence was not listed. Each offence was assigned an offence severity score based on the Youth Justice Board Counting Rules ranging from 1 (e.g. minor public order offences) to 8 (e.g. murder). These data were collected with the intention of identifying subgroups within the offender group for further analysis.

Multidimensional Scale of Perceived Social Support

Participants were requested to complete the Multidimensional Scale of Perceived Social Support (MSPSS; Zimet et al., 1988). The MSPSS is the most widely used measure of perceived social support (Osman et al., 2014) and has demonstrated positive psychometric properties among adolescents, with good levels of internal consistency and construct reliability (Bruwer et al., 2008). The 12-item self-report measure assesses perceived availability and adequacy of emotional and instrumental support. Each item is rated on a 7-point Likert-type scale ranging from ‘very strongly disagree’ to ‘very strongly agree’. Total subscale scores (family, friends, significant other) range from 4 to 28 and total composite MSPSS scores range from 12 to 84, with a higher score indicating higher levels of perceived social support.

Toronto Alexithymia Scale

Participants were requested to complete the Toronto Alexithymia Scale (TAS-20; Bagby et al., 1994a, 1994b) which is considered a robust measure of alexithymia (Karukivi et al., 2011). The TAS-20 is a 20-item, self-report measure that uses a 5-point Likert-type scale, asking participants the extent to which they believe each statement applies to them, ranging from ‘strongly disagree’ to ‘strongly agree’. The TAS-20 provides an overall alexithymia score, ranging from 20 to 100, as well as three inter-correlated subscale scores: difficulty identifying feelings (DIF), difficulty describing feelings (DDF) and externally oriented thinking (EOT).

TAS-20 psychometric properties have been confirmed with samples of young people (Parker et al., 2010). It has been recommended that adaptations are made to alleviate reading deficits and co-morbid difficulties, such as inattention (Parker et al., 2010). Therefore, all participants wore headphones throughout questionnaire completion to listen to the audio- recorded TAS-20 statements alongside the written statements, and the questionnaire was computer-administered with the aim of enhancing attention levels.

FER

Designed and administered through the Medialab application (Jarvis, 2012), participants were requested to complete the FER task, as developed by Bowen et al. (2013) and based on Ekman and Friesen’s (1976) facial affect battery. The measure has good reliability (Frank and Stennett, 2001). Research has been completed with children and adolescents using Ekman and Friesen’s (1976) photographs and morphing these with neutral expressions to create different levels of emotion intensities (Montirosso et al., 2010).

Aiming to minimise task fatigue alongside administration of the other measures, a briefer version of the FER task was used (with permission of the developer). Participants were shown the 34 facial expressions in random order, consisting of 2 neutral expressions (1 male and 1 female photo) and 2 repeats of four emotional intensities (25%, 50%, 75%, and 100% by being morphed with their matching neutral expression), for four emotional categories (happiness, sadness, fear and anger) (16 male and 16 female photos). The question, ‘What emotion is this person showing?’ accompanied each item, and participants were required to indicate their response by selecting an emotional category (listed in the same order each time). No time limits were applied.

VEPR

Participants were also requested to complete the VEPR Task (Davies, 2015), which, when piloted, reached good reliability (Davies, 2015). Participants were asked to listen to 30 randomly presented, lexically neutral statements ‘his glasses are on the table’ (Boaz et al., 2011), spoken with neutral prosody and four emotional tones (anger, fear, happiness and sadness), each repeated 6 times (15 spoken by a male actor and 15 spoken by a female actor). The VEPR was piloted, with all items reaching good reliability (r > 0.80). The question, ‘What is the speaker’s emotion?’ accompanied each item and participants were required to indicate their response by selecting an emotional category (listed in the same order each time). No time limits were applied and audio statements were played only once.

Statistical methods

Preliminary analysis revealed that several variables violated the assumptions required for parametric data analysis, so bootstrapping methods (Efron and Tibshirani, 1993) were used as a robust approach to statistical analyses (Field, 2013). Bootstrapping methods of 2000 samples were used for t-test and ANOVA analyses, allowing inferences to be made on normally and non-normally distributed data (Field, 2013; Wright et al., 2011).

Descriptive statistics were used to present the demographic characteristics of the sample and to collate basic findings. Preliminary analysis identified significant between group differences relating to several demographic factors. Bootstrapped t-tests were carried out to determine whether these variables would significantly confound group differences in measures of emotion recognition and perceived support. Correlational analyses were used to explore the relationship between the variables measured (emotion recognition and social support). ANOVA and ANCOVA analyses were completed to establish between group differences in TAS, VEPR and MSPSS total and subscale scores. Repeated measures MANOVA analyses of FER scores were completed to establish effects of emotion intensity, effects of group and interaction effects between group and emotion intensity. Lastly, bootstrapped two-tailed t-tests and repeated measures MANOVA tests were completed to identify any between group differences in the young people who offend subgroups identified.

Results

Demographic characteristics and relationship to outcome variable

Statistical analyses revealed no significant difference between groups in terms of age, gender, grades, ethnicity (collapsing ‘non-White British’ ethnicities), qualifications and SES (p > 0.05). However, bootstrapped t-tests revealed that several demographic variables were significantly related to emotion recognition and perceived support (and, as confounding variables, were controlled for in subsequent analysis).

Relative to females, males reported significantly lower levels of alexithymia (TAS-20; TAS-DIF; TAS-DDF) and lower FER scores of happiness (total and 100%), sadness (at 50% intensity) and anger (at 75% intensity). Compared with young people with higher academic grades (A*-C), young people with lower grades (D-fails) reported significantly more difficulty describing feelings and gained significantly lower scores of emotion recognition through verbal prosody (total and happy) and facial expressions (happy at 75% and 100% intensity, sadness at 100% intensity and anger at 100% intensity).

Employment status differed significantly between the groups (= 0.42, p = 0.01), with 72% of the controls and 44% of Young people who offend studying and 6% of the controls and 26% of young people who offend working. Accommodation status differed significantly between the groups (χ² = 14.91, p < 0.001), with 96% of the controls and 64% of the young people who offend living with their family or partner. Accommodation status was associated with VEPR of anger, FER of fear (50% intensity) and anger (100% intensity), and perceived social support (total and family subscale). Therapeutic input differed significantly between the groups (χ² = 8.76, p < 0.01), with 40% of the young people who offend and 14% of the controls reporting that they had received therapeutic input. Therapy status was associated with alexithymia and levels of perceived support from family.

Care status differed significantly for the two groups (χ² = 15.43, p < 0.001), with 38% of young people who offend and 4% of the controls having spent time in care. Of particular interest, having spent time in care was significantly associated to all measures: ability to identify feelings (t (98) = 2.07, p < 0.05) and describe feelings (t (98) = 2.00, p < 0.05), recognition of all emotions through verbal prosody (t (98) = −2.61, p < 0.05), FER of sadness (at 50% intensity) (t (98) = −2.13, p < 0.05) and perceived support (t (98) = −2.33, p < 0.05), especially from family (t (98) = −3.76, p < 0.01).

Relationship between variables

Whole group analysis (N = 100) found a significant negative correlation between TAS-20 scores and MSPSS total scores (p < 0.05, r = −0.19) and significant positive correlations between VEPR total and MSPSS total scores (p < 0.05), FER total and MSPSS total scores (p < 0.01) and FER total and VEPR total scores (p < 0.001, r = 0.66). Significant correlations were not found between VEPR total and TAS-20 total scores and FER total and TAS-20 total scores.

Between-group analysis

Alexithymia

Failing to support the hypothesis, bootstrapped univariate ANCOVA analyses (controlling for therapeutic input for TAS-20, TAS-DIF and TAS-DDF and Care status for TAS-DIF and TAS-DDF), revealed no significant differences between groups for TAS-20, TAS-DIF or TAS-DDF scores.

VEPR

Failing to support the hypothesis, bootstrapped univariate ANOVA tests and ANCOVA tests (controlling for Care status for VEPR total and VEPR fear; and accommodation for VEPR anger), revealed no significant group differences, with the exception of VEPR of fear (F(1,97) = 4.54, p < 0.05, η_p² = 0.05). Inspection of group means revealed significantly lower accuracy of VEPR of fear for the young people who offend group (M = 55.67, SD = 21.72), than for the controls (M = 69.33, SD = 23.17). It is worth noting that, relative to the other emotions, VEPR of fear had markedly lower scores across both groups.

FER

For happiness, there was a main effect of intensity (F(2.29,225.13) = 179.68, p < 0.01), no main effect of group (F(1,98) = .19, p = 0.66) and no significant interaction between intensity and group (F(2.29,225.13) = 0.95, p = 0.39).

For sadness, there was a main effect of intensity (F(2.72,266.74) = 75.33, p < 0.01), no main effect of group (F(1,98) = 1.98, p = 0.16) and no significant interaction between intensity and group (F(2.72,266.74) = 2.12, p = 0.10).

For fear, there was a main effect of intensity (F(2.48, 242.58) = 56.53, p < 0.001), no main effect of group (F (1,98) = 2.45, p = 0.12) and no significant interaction between intensity and group (F(2.48, 242.58) = 0.23, p = 0.84).

For anger, there was a main effect of intensity (F(2.51, 245.46) = 125.51, p < 0.001), no main effect of group (F(1,98) = 2.21, p = 0.14) and no significant interaction between intensity and group (F (2.51, 245.46) = 0.83, p = 0.46).

A significant between group difference was found for neutral recognition (F(1,98) = 17.50, p < 0.001). Inspection of group means revealed significantly lower accuracy of neutral FER within the young people who offend (M = 65.00, SD 30.72), than the controls (M = 88.00, SD = 88.00).

Perceived social support

Failing to support the hypothesis, ANCOVA analyses (controlling for looked after in care status for MSPSS total and MSPSS family and therapy status for MSPSS family) revealed no significant group differences.

Within the young people who offend

Within young people who offend group analysis only revealed significant findings in the offence type group (violent compared to non-violent offenders). Within young people who offend group analysis according to offence severity (low offence severity ⩽4 and high offence severity ⩾5) and offence frequency (⩽3 and ⩾4 offences), showed no significant findings. Relevant between group analyses described for each subgroup should be interpreted with caution due to small subgroup sample sizes.

Offence type

In terms of alexithymia, bootstrapped ANOVA analysis (and ANCOVA analysis for TAS-total, TAS-DIF and TAS-DDF scores) revealed significant differences between the young people who offend subgroups in TAS-total scores (F(1,45) = 4.09, p < 0.05,η_p² = 0.08) and TAS-EOT scores (F(1,46) = 4.69, p < 0.05,η_p² = 0.09). Inspection of group means revealed that, relative to ‘non-violent only’ young people who offend, ‘violent’ young people who offend reported lower TAS-20 scores and lower TAS-EOT scores. Bootstrapped two-tailed t-tests revealed that, relative to ‘non-violent only’ young people who offend, ‘violent’ young people who offend reported significantly higher MSPSS friend scores (equal variances not assumed), t(46) = 2.47, p < 0.05) and significantly lower MSPSS significant other scores (t(46) = −2.35, p < 0.05).

Discussion

The current study is the first known study to examine emotion recognition through verbal prosody and both emotion recognition and perceived social support with young people who offend, including males and females, while controlling for potentially confounding variables and completing within group analysis of young people who offend. The primary aim of this study was to examine whether, relative to controls, young people who offend would (1) show significantly lower levels of emotion recognition and (2) lower levels of perceived social support. However, higher alexithymia scores and lower FER, VEPR and MSPSS scores for the young people who offend, compared to controls, did not reach significance (with the exception of FER of neutral and VEPR of fear). Further aims of the study were to examine correlations between all outcomes measures and, indeed, significant relationships were found between most variables.

Emotion recognition

Alexithymia

Higher scores in TAS-20 total, DDF and DIF subscales among the young people who offend group did not reach significance. This is consistent with some (Möller et al. 2014; Moriarty et al., 2001), but not all (Zimmermann, 2006), previous research. It should be noted that initial t-tests (with and without outliers) conducted with the current sample, indicated that young people who offend showed significantly lower DIF and DDF scores. The discrepancy between current study findings and some previous research might therefore be due to confounding variables, such as looked after in care status (which was significantly related to difficulty identifying and describing feelings); a variable not previously controlled for.

Subgroup analysis only showed significant between-group differences in relation to offence type, whereby ‘violent’ young people who offend showed significantly lower levels of alexithymia overall and externally oriented thinking, than ‘non-violent only’ young people who offend. These findings are unexpected, as alexithymia has been associated with a reduced ability to regulate one’s emotions, increasing the risk of violent expression of emotional states (Fossati et al., 2009). Furthermore, these findings contradict previous research by Möller et al. (2014), who reported that, although not statistically significant, young people with violent offences scored higher on TAS-20 total and subscale scores. Of note, Möller et al. (2014) recruited older participants (aged 18–21) from prison, did not control for confounding variables and categorised participants as violent or non-violent based on the principal offence (as opposed to any violent offence).

VEPR

Partially supporting the hypothesis, young people who offend showed significantly lower accuracy in verbal emotional prosody of fear. No previous studies are known to have examined emotion recognition through verbal prosody with a non-clinical adolescent or young people who offend sample, so there is little relevant research with which to compare these results. The findings are consistent with studies including samples with similar social, emotional and cognitive regulation difficulties and adverse early relationships and events, including adult forensic samples (Bagley et al., 2009; Suchy et al., 2009) and pupils attending schools which support social, emotional and behavioural needs (Blair et al., 2005; Dawel et al., 2012).

These study findings are supportive of previous research proposing that a reduced ability to recognise negative emotions (particularly fear), can lead to reduced feelings of punishment by others’ fearful expressions, reducing the likeliness of an empathic response and inhibition of the behaviour that caused this distress (Bird and Viding, 2014; Blair, 2005; Murray and Andrews, 2005).

FER

Failing to support the main hypothesis, no significant differences were found between young people who offend and the controls in total scores for FER of happiness, sadness, fear or anger. Although these findings are consistent with some previous research (Carr and Lutjemeier, 2005; Gonzalez-Gadea et al., 2014; Jones et al., 2007; McCown et al., 1988; Sato et al., 2009), previous research using FER measures more realistic to everyday situations (like the current study), reported that young people who offend do show significantly lower FER scores than controls (Bowen et al., 2013; Gonzalez-Gadea et al., 2014). Confounding variables might be responsible for the discrepancy between research findings. For example, initial t-test analysis of the current sample showed significant between group differences in overall FER scores, with (p < 0.01) and without outliers (p < 0.02), but these differences no longer remained significant when controlling for looked after in care and accommodation status, factors not controlled for in previous research.

A significant between-group difference was found in FER of neutral expression (p < 0.001), with young people who offend scoring significantly lower than controls. Inaccurate recognition of neutral facial expressions were attributable to an incorrect selection of sadness. This might be supportive of negative attribution theory, whereby young people who offend are more likely to make negative interpretations of another’s emotions and intent (Dodge, 2006), although this attribution was found in both participant groups.

Perceived social support

In support of previous research revealing high levels of social support needs of young people who offend (Chitsabesan et al., 2006; King et al., 2014), findings from the current study showed that young people who offend reported lower mean scores of perceived social support than controls. However, failing to support the hypothesis, these differences did not reach significance, which might be due to young people who offend, more so than controls, wanting to present as socially acceptable (Hardan-Khalil and Mayo, 2015). Subgroup analysis in the young people who offend sample also revealed no significant between-group differences in levels of perceived social support.

In relation to offence type, relative to ‘non-violent only’ young people who offend, ‘violent’ young people who offend reported higher levels of social support from friends, but lower levels of social support from a significant other person. Indeed, previous research suggests that young people who offend appear to experience a number of personal and sociocultural barriers to using social support for positive outcomes (King et al., 2014). It is generally understood that adolescents turn to peers rather than accessing support from family or adults (De Goede et al., 2009; Lambourn, 2009). However, these findings are supportive of previous research findings that social supportive friendships may be not always have a positive influence, rather increase the risk of offending, by assisting young people in gaining knowledge, skills, connections, role models, a sense of belonging and social status (Cullen, 1994; Salvatore and Markowitz, 2014).

Looked after in care status

Of particular interest, analysis showed that looked after in care status, more commonly reported in the young people who offend sample (38%) than control sample (4%), was the predominant factor significantly associated with all measures of emotion recognition and perceived social support. Specifically, and in line with previous research (Barone and Lionetti, 2012; Hollingworth, 2014), compared with young people without looked after in care status, young people with looked after in care status showed reduced ability to identify and describe feelings and reduced FER of sadness (at 50% intensity) and verbal emotional prosody (particularly fear). These findings also support research findings that early adversity and socio-economic deprivation (likely to be experienced by looked after in care status) are related to emotion recognition deficits, especially recognition of negative emotions (Esposito et al., 2014; Fox et al., 2010; Murray and Andrews, 2005).

Significantly lower levels of social support (particularly from family) among young people with looked after in care status provides further support for studies reporting that attachment experiences, such as caregiver availability, responsiveness and acceptance shape an individual’s sense of acceptance, appraisal and attention to, and availability of, social support (Lakey and Cohen, 2000; Ma, 2006; Mikulincer et al., 2003).

Relationship between emotion recognition and perceived social support

Consistent with previous research (Mill et al., 2009; Rigoulot and Pell, 2014), a strong association was found between emotion recognition from facial expressions and verbal prosody. These findings are also supportive of neurological research positing that paralinguistic communication is primarily processed in the same brain regions (Mill et al., 2009).

Also as hypothesised, emotion recognition (alexithymia, FER and verbal emotional prosody) was found to be related to perceived social support. This is consistent with previous research reporting that emotion recognition (of oneself and others) plays an important role in social interaction and relationships (Collin et al., 2013; Karukivi et al., 2014; Spitzer et al., 2005).

Failing to support the hypothesis and previous research (Israelashvili et al., 2019; Jongen et al., 2014), analysis showed no significant association between ability to recognise emotions in oneself and others. However, unlike the current study, previous research includes time restraints, which is thought to support individuals with alexithymia to correctly label others’ emotions by relying on information related to the visual configuration of the facial expression rather than affective cues (Jongen et al., 2014). Furthermore, all previous research includes adults and some categorised participants into groups of alexithymics and non-alexithymics, which is likely to have affected significance of results.

Clinical implications

Emotion recognition interventions have been shown to improve FER (Van Goozen et al., 2013) and social relationships, and to reduce offending behaviour (Dadds et al., 2012; Hubble et al., 2015). However, it is recommended that relation-based interventions should be provided first (Skuse and Matthew, 2015), such as those based on the trauma-recovery model (TRM), which support young people who offend to manage the impact of early adversity and developmental trauma by meeting basic needs and building relationships before skills building (Skuse and Matthew, 2015). Relation-based therapies such as dialectical behaviour therapy (Linehan, 1993) and dyadic developmental psychotherapy (Hughes, 2006) are recommended for young people with looked after in care status (Andrew et al., 2014; Hughes et al., 2015; Quinn and Shera, 2009).

In terms of social support, prior to simply offering support-increasing interventions, there is a need for individualised assessment to consider the complex interactions of mediating factors of support benefits (Martinez and Abrams, 2013), as evidenced by the findings that young people who offend with violent offences reported significantly higher levels of perceived social support from friends. Consideration should also be given to the bidirectional relationship between a young person’s difficulties and their surrounding system, focusing intervention throughout a young person’s entire ecosystem (Phoenix, 2016; Preston et al., 2015).

Conclusion

Policy and legislation in relation to young people who offend emphasise the need to do more to understand the complex interplay of factors leading a young person to offend and re-offend. Accordingly, the current study established that offending behaviour is likely to be the outcome of a complex interplay of individual, developmental, and social factors. Developmental theoretical frameworks suggest that early adversity can predispose emotional skills deficits which reduce one’s ability to use socially acceptable ways to express and regulate feelings, such as aggressive impulses, setting the stage for offending behaviour. Indeed, the current study indicated that looked after in care status (representative of early adversity), was more commonly reported among young people with, than young people without, an offending history. Furthermore, looked after in care status, rather than offending status in isolation, was found to be significantly related to lower levels of perceived social support and difficulties recognising emotions in oneself and others. Thus, the study suggests that cognitive, emotional and social functioning is likely to mediate the link between early adverse experiences and behaviour problems (Van Goozen et al., 2007). Accordingly, services and professionals need to ‘work together to change perceptions of young people who offend. . . to better understand the needs of these. . .vulnerable young people and how their self-belief, skills and achievements can be encouraged to give them better chances in life’ (Welsh Government/Youth Justice Board, 2014: 2).

Footnotes

Acknowledgements

The authors particularly thank the young people and staff at South East Wales Youth Offending Teams, Coleg Gwent and Torfaen Youth Service, who made this research possible. They also thank Professor Stephanie van Goozen for supporting the initial development of the research idea and providing the Facial Emotion Recognition task, and Sue Thomas for her guidance in getting the idea off the ground. They are grateful to Dr Holly Davies for allowing to use the Verbal Emotional Prosody Task she developed.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Maria Wilcox

Author biographies

Maria Wilcox completed her undergraduate Psychology Degree in the University of Glamorgan and worked in specialist services for adults with a learning disability for 5 years before completing her Doctorate in Clinical Psychology in Cardiff University in 2016. Throughout training and post-qualification, Maria has maintained interest in service development and delivery of attachment and trauma-informed children’s services, specifically services for looked after children and children with additional needs. Maria is currently Clinical Lead for an Intensive Positive Behaviour Support Service for children and young people with a learning disability and/or autism in Aneurin Bevan University Health Board.

Neil Frude trained at the Institute of Psychiatry in London and then obtained a PhD in the University of Wales. For the past 40 years he has been engaged in therapy, training and research. Neil is currently a Consultant Clinical Psychologist and an Honorary Professor in Cardiff University and the University of South Wales. He has published a number of books on topics as diverse as family relations, disruption in schools, violence and human interaction with computers. In 2003, Neil devised a book prescription scheme for mental health which has since become a national scheme in Wales and England as well as being introduced in several other countries. Neil is a Fellow of the British Psychological Society and was recently given a Lifetime Achievement Award by the Society. He has also performed as a stand-up comedian at the Edinburgh Fringe Festival.

Liz Andrew graduated from Swansea University in 1998 with a First Class Honours Degree in Psychology and spent some time working in Psychological Medicine at the Institute of Psychiatry before returning to Wales in 2000 to complete her Doctorate in Clinical Psychology. In 2009, Liz completed a research fellowship at Cardiff’s School of Psychology funded by the Institute of Forensic Mental Health. This culminated in her obtaining a PhD in Forensic Psychology. Liz is a Consultant Clinical Psychologist and Head of Adult Psychology in Aneurin Bevan University Health Board. Liz has 16 years’ experience of working with individuals with complex psychological needs and severe levels of emotional distress and has led the development of innovative interventions for individuals at high risk of self-harm and suicide. Liz maintains research interests in the relationship between attachment and psychological distress, alexythymia, trauma and outcomes from evidence based care.

References

Andrew

Williams

Waters

(2014) Dialectical behaviour therapy and attachment: Vehicles for the development of resilience in young people leaving the care system. Clinical Child Psychology and Psychiatry 19(4): 503–515.

Andrews

James

(2010) Rehabilitating criminal justice policy and practice. Psychology, Public Policy and Law 16(1): 39–55.

Bagby

Parker

Taylor

(1994a) The twenty-item Toronto Alexithymia Scale: I. Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research 38: 23–32.

Bagby

Taylor

Parker

JDA

(1994b) The twenty-item Toronto Alexithymia Scale: II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic Research 38: 33–40.

Bagley

Abramowitz

Kosson

(2009) Vocal affect recognition and psychopathy: Converging findings across traditional and cluster analytic approaches to assessing the construct. Journal of Abnormal Psychology 118(2): 388–398.

Bagnall

South

Hulme

Woodall

Vinall-Collier

Raine

Kinsella

Dixey

Harris

Wright

(2015) A systematic review of the effectiveness and cost-effectiveness of peer education and peer support in prisons. BMC Public Health 15(1): 290–320.

Barone

Lionetti

(2012) Attachment and emotional understanding: A study on late-adopted pre-schoolers and their parents. Child, Care, Health and Development 3(5): 690–696.

Barrera

(1986) Distinctions between social support concepts, measures and models. American Journal of Community Psychology 14: 413–445.

Bartholomew

Cobb

Poole

(1997) Adult attachment patterns and social support processes. In: Pierce

Lakey

Sarason

(eds) Sourcebook of Social Support and Personality. London: Plenum Press, 359–378.

10.

Berastegui

Van Leeuwen

Chabrol

(2012) Relationships between emotional intelligence, alexithymia and interpersonal delinquent behaviour in a sample of high school students. Encaphale 38(5): 426–432.

11.

Bergseth

Bouffard

(2013) Examining the effectiveness of a restorative justice program for various types of juvenile offenders. International Journal of Offender Therapy and Comparative Criminology 57: 1054–1075.

12.

Biehal

Ellison

Baker

Sinclair

(2010) Belonging and Permanence: Outcomes in Long-Term Foster Care and Adoption. London: BAAF.

13.

Bird

Viding

(2014) The self to other model of empathy: Providing a new framework for understanding empathy impairments in psychopathy, autism and alexithymia. Neuroscience and Biobehavioral Reviews 42: 520–532.

14.

Blades

Hart

Lea

Willmott

(2011) Care – A Stepping Stone to Custody? London: Prison Reform Trust.

15.

Blair

RJR

(2005) Applying a cognitive neuroscience perspective to the disorder of psychopathy. Development and Psychopathology 17: 865–891.

16.

Blair

RJR

Budhani

Colledge

Scott

(2005) Deafness to fear in boys with psychopathic tendencies. Journal of Child Psychology and Psychiatry 46(3): 327–336.

17.

Blair

RJR

Colledge

Murray

Mitchell

(2001) A selective impairment in the processing of sad and fearful expressions in children with psychopathic tendencies. Journal of Abnormal Child Psychology 29: 491–498.

18.

Boaz

MB-D

Van Lieshout

PHHM

Leszcz

(2011) A resource of validated affective and neutral sentences to assess identification of emotion in spoken language after a brain injury. Brain Injury 2011(2): 25206–25220.

19.

Bowen

Morgan

Moore

Van Goozen

SHM

(2013) Young offenders’ emotion recognition dysfunction across emotion intensities: Explaining variation using psychopathic traits, conduct disorder and offence severity. Journal of Psychology and Behavioural Assessment 36(1): 60–73.

20.

Bruwer

Emsley

Kidd

Lochner

Seedat

(2008) Psychometric properties of the Multidimensional Scale of Perceived Social Support in youth. Comprehensive Psychiatry 49: 195–201.

21.

Burleson

MacGeorge

(2002) Supportive communication. In: Knapp

Daly

(eds) Handbook of Interpersonal Communication, 3rd Edition. Thousand Oaks, CA: Sage, 374–422.

22.

Carr

Lutjemeier

(2005) The relation of facial affect recognition and empathy to delinquency in youth offenders. Adolescence 40(159): 601–619.

23.

Chen

Feeley

(2012) Enacted support and well-being: A test of the mediating role of perceived control. Communication Studies 63: 608–625.

24.

Chitsabesan

Kroll

Bailey

Kenning

Sneider

MacDonald

Theodosiou

(2006) Mental health needs of young offenders in custody and in the community. The British Journal of Psychiatry 188(6): 534–540.

25.

Clear

Rose

Ryder

(2001) Incarceration and the community: The problem of removing and returning offenders. Crime & Delinquency 47: 335–351.

26.

Collin

Bindra

Raju

Gillberg

Minnis

(2013) Facial emotion recognition in child psychiatry: A systematic review. Research in Developmental Disabilities 34(5): 1505–1520.

27.

Colvin

Cullen

Vander Ven

(2002) Coercion, social support, and crime: An emerging theoretical consensus. Criminology 40(1): 19–42.

28.

Cullen

(1994) Social support as an organizing concept for criminology: Presidential address to the Academy for criminal justice sciences. Justice Quarterly 11: 528–559.

29.

Dadds

Cauchi

Wimalaweera

Hawes

Brennan

(2012) Outcomes, moderators, and mediators of empathic-emotion recognition training for complex conduct problems in childhood. Psychiatry Research 199(3): 201–207.

30.

Dadds

Perry

Hawes

Merz

Riddell

Haines

Solak

Abeygunawardane

(2006) Attention to the eyes and fear-recognition deficits in child psychopathy. British Journal of Psychiatry 189: 280–281.

31.

Davies

(2015) Social cognition and interpersonal relationships in individuals with post-traumatic stress disorder (PTSD). Unpublished Doctoral Thesis, Cardiff University, Cardiff.

32.

Dawel

O’Kearney

McKone

Palermo

(2012) Not just fear and sadness: Meta-analytic evidence of pervasive emotion recognition deficits for facial and vocal expressions in psychopathy. Neuroscience and Biobehavioural Reviews 36: 2288–2304.

33.

De Goede

IHA

Branje

SJT

Delsing

MJMH

Meeus

WHJ

(2009) Linkages over time between adolescents’ relationships with parents and friends. Journal of Youth and Adolescence 38(10): 1304–1315.

34.

Dodge

(2006) Translational science in action: Hostile attributional style and the development of aggressive behavior problems. Developmental Psychopathology 18(3): 791–814.

35.

Efron

Tibshirani

(1993) An Introduction to the Bootstrap. New York: Chapman & Hall.

36.

Ekman

Friesen

(1976) Pictures of Facial Affect. Palo Alto, CA: Consulting Psychologists Press.

37.

Esposito

Palumbo

Troncone

(2014) The influence of the attachment style on the decoding accuracy of emotional vocal expressions. Cognitive Computations 6(4): 699–707.

38.

Fairchild

Van Goozen

SHM

Calder

Stollery

Goodyer

(2009) Deficits in facial expression recognition in male adolescents with early-onset or adolescent-onset conduct disorder. Journal of Child Psychology and Psychiatry 50: 627–636.

39.

Field

(2013) Discovering Statistics Using IBM SPSS Statistics, 4th Edition. London: Sage.

40.

Fitness

(2006) The emotionally intelligent marriage. In: Ciarrochi

Forgas

Mayer

(eds) Emotional Intelligence in Everyday Life, 2nd Edition. New York: Psychology Press; Taylor & Francis, 129–139.

41.

Fossati

Acquarini

Feeney

Borroni

Grazioli

Giarolli

Franciosi

Maffei

(2009) Alexithymia and attachment insecurities in impulsive aggression. Attachment and Human Development 11(2): 165–182.

42.

Fox

Levitt

Nelson

III (2010) How the timing and quality of early experiences influence the development of brain architecture. Child Development, 2010(1): 8128–8140.

43.

Frank

Stennett

(2001) The forced-choice paradigm and the perception of facial expressions of emotion. Journal of Personality and Social Psychology 80: 70–85.

44.

Gonzalez-Gadea

Herrera

Parra

Mendez

Baez

Manes

Ibanez

(2014) Emotion recognition and cognitive empathy deficits in adolescent offenders revealed by context-sensitive tasks. Frontiers in Human Neuroscience 8: 1–11.

45.

Grynberg

Chang

Corneille

Maurage

Vermeulen

Berthoz

Luminet

(2012) Alexithymia and the processing of emotional facial expressions (EFEs): Systematic review, unanswered questions and further perspectives. PLoS ONE 7(8): e42429.

46.

Hardan-Khalil

Mayo

(2015) Psychometric properties of the Multidimensional Scale of Perceived Social Support. Clinical Nurse Specialist 29(5): 258–261.

47.

Hollingworth

(2014) Social cognition, attachment and emotional regulation in young adults leaving care. Unpublished dissertation, Cardiff University, Cardiff.

48.

Hubble

Bowen

Moore

van Goozen

SHM

(2015) Improving negative emotion recognition in young offenders reduces subsequent crime. PLoS ONE 10: e0132035.

49.

Hughes

Golding

Hudson

(2015) Dyadic developmental psychotherapy (DDP): The development of the theory, practice and research base. Adoption & Fostering 39(4): 356–365.

50.

Hughes

(2006) Building the Bonds of Attachment: Awakening Love in Deeply Troubled Children. Oxford: Rowman & Littlefield Publishers.

51.

Israelashvili

Oosterwijk

Sauter

Fischer

(2019) Knowing me, knowing you: Emotion differentiation in oneself is associated with recognition of others’ emotions. Cognition and Emotion 33: 1461–1471.

52.

James

Stams

GJJM

Asscher

De Roo

Van der Laan

(2013) Aftercare programs for reducing recidivism among juvenile and young adult offenders: A meta-analytic review. Clinical Psychology Review 33(2): 263–274.

53.

Jarvis

(2012) MediaLab (Version 2012, Computer Software). New York: Empirisoft Corporation.

54.

Johnson

Esposito-Smythers

Miranda

Jr Rizzo

Justus

Clum

(2011) Gender, social support and depression in criminal justice-involved adolescents. International Journal of Offender Therapy and Comparative Criminology 55(7): 1096–1109.

55.

Jones

Forster

Skuse

(2007) What do you think you’re looking at? Investigating social cognition in young offenders. Criminal Behaviour and Mental Health 17: 101–106.

56.

Jongen

Axmacher

Kremers

NAW

Hoffmann

Limbrecht-Ecklundt

Traue

Kessler

(2014) An investigation of facial emotion recognition impairments in alexithymia and its neural correlates. Behavioural Brain Research 271: 129–139.

57.

Jusyte

Mayer

Künzel

Hautzinger

Schönenberg

(2014) Unemotional traits predict early processing deficit for fearful expressions in young violent offenders: An investigation using continuous flash suppression. Psychological Medicine 45(2): 285–297.

58.

Karukivi

Joukamaa

Hautala

Kaleva

Haapasalo-Pesu

Liuksila

Saarijarvi

(2011) Does perceived social support and parental attitude relate to alexithymia? A study in Finnish late adolescents. Psychiatry Research 187: 254–260.

59.

Karukivi

Pölönen

Vahlberg

Saikkonen

Saarijärvi

(2014) Stability of alexithymia in late adolescence: Results of a 4-year follow-up study. Psychiatry Research 219: 386–390.

60.

King

Brown

Peth

Wright

(2014) Perceptions of support-seeking in young people attending a Youth Offending Team: An interpretative phenomenological analysis. Clinical Child Psychology and Psychiatry 19(7): 7–23.

61.

Lakey

Cohen

(2000) Social support theory and measurement. In: Cohen

Underwood

Gottlieb

(eds) Social Support Measurement and Intervention: A Guide for Health and Social Scientists. New York: Oxford University Press, 29–52.

62.

Lambourn

(2009) Factors Affecting Young People’s and Adolescents’ Willingness to Seek Help for Mental Health Problems. Unpublished Master’s thesis, University of Westminster, London.

63.

Langevin

Hare

(2001) Psychopathie et alexithymie chez un groupe de jeunes contrevenants. Revue de Psychoeducation et d’Orientation 30(2): 227–236.

64.

Leach

(2015) Improving Mental Health through Social Support: Building Positive and Empowering Relationships. London: Jessica Kingsley Publishers.

65.

Linehan

(1993) Cognitive-Behavioral Treatment of Borderline Personality Disorder. New York: Guilford Press.

66.

López

Cooper

(2011) Social Support Measures Review. National Centre for Latino Child & Family Research. Available at: https://www.first5la.org/files/SSMS_LopezCooper_LiteratureReviewandTable_02212011.pdf

67.

(2006) Attachment theory in adult psychiatry. Part 1: Conceptualisations, measurement and clinical research findings. Advances in Psychiatric Treatment 12: 440–449.

68.

McCown

Johnson

Austin

(1986) Inability of delinquents to recognize facial affects. Journal of Social Behavior and Personality 1(4): 489–496.

69.

McCown

Johnson

Austin

(1988) Patterns of facial affect recognition errors in delinquent adolescent males. Journal of Social Behavior and Personality 3(3): 215–224.

70.

Marsh

Blair

(2008) Deficits in facial affect recognition among antisocial populations: A meta-analysis. Neuroscience and Biobehavioural Reviews 32: 454–465.

71.

Martinez

Abrams

(2013) Informal social support among returning young offenders: A meta-synthesis of the literature. International Journal of Offender Therapy and Comparative Criminology 57(2): 169–190.

72.

Meins

Fernyhough

Wainwright

Das Gupta

Fradley

Tuckey

(2002) Maternal mind-mindedness and attachment security as predictors of theory of mind understanding. Child Development 73(6): 1715–1726.

73.

Mikulincer

Shaver

Pereg

(2003) Attachment theory and affect regulation: The dynamics, development, and cognitive consequences of attachment-related strategies. Motivation and Emotion 27: 77–102.

74.

Mill

Allik

Realo

Valk

(2009) Age related differences in emotion recognition: A cross-sectional study. Emotion 9(5): 619–630.

75.

Ministry of Justice (MoJ) (2008) Factors linked to re-offending: A one year follow-up of prisoners who took part in the Resettlement Surveys 2001, 2003, 2004 (MoJ Research Summary 5). Available at: http://webarchive.nationalarchives.gov.uk/20110201125714/http://www.justice.gov.uk/publications/docs/research-factors-reoffending.pdf (accessed 5 December 2015).

76.

Ministry of Justice and Department for Children, Schools and Families (2009) Reducing re-offending: Supporting families, creating better futures – A framework for improving the local delivery of support for the families of offenders. Available at: http://dera.ioe.ac.uk/207/7/reducing-reoffending-supporting-families_Redacted.pdf (accessed 5 December 2015).

77.

Möller

Falkenström

Larsson

Holmqvist

(2014) Mentalizing in young offenders. Psychoanalytic Psychology 31(1): 84–99.

78.

Monahan

Goldweber

Cauffman

(2011) The effects of visitation on incarcerated Juvenile offenders: How contact with the outside impacts adjustment on the inside. Law & Human Behavior 35(2): 143–151.

79.

Montirosso

Peverelli

Frigerio

Crespi

Borgatti

(2010) The development of dynamic facial expression recognition at different intensities in 4- to 18-year-olds. Social Development 19(1): 71–92.

80.

Moriarty

Stough

Tidmarsh

Eger

Dennison

(2001) Deficits in emotional intelligence underlying adolescent sex offending. Journal of Adolescence 24: 743–751.

81.

Murray

Andrews

(2005) The Social Baby: Understanding Babies’ Communication from Birth. London: CP Publishing.

82.

Osman

Lamis

Freedenthal

Gutierrez

McNaughton-Caddill

(2014) The Multidimensional Scale of Perceived Social support: Analyses of internal reliability, measurement invariance and correlates across gender. Journal of Personality Assessment 96(1): 103–112.

83.

Parker

Eastabrook

Keefer

Wood

(2010) Can alexithymia be assessed in adolescents? Psychometric properties of the 20-item Toronto Alexithymia Scale in younger, middle and older adolescents. Psychological Assessment 2(4): 798–808.

84.

Phoenix

(2016) Against youth justice and youth governance, for youth penality. The British Journal of Criminology 56: 123–140.

85.

Preston

Lister

Shostal

(2015) What good looks like in psychological services for children, young people and their families. Paper 10: Delivering psychological services for children and young people involved with the criminal justice system, those are risk of involvement and their families. Child and Family Clinical Psychology Review 3: 130–140.

86.

Quinn

Shera

(2009) Evidence-based practice in group work with incarcerated youth. International Journal of Law and Psychiatry 32(5): 288–293.

87.

Rigoulot

Pell

(2014) Emotion in the voice influence the way we scan emotional faces. Speech Communication 65: 36–49.

88.

Salovey

Mayer

(1990) Emotional intelligence. Imagination, Cognition and Personality 9: 185–211.

89.

Salvatore

Markowitz

(2014) Do life course transitions and social bonds influence male and female offending differently? Gender Contrasts and Criminality, Deviant Behavior 35(8): 628–653.

90.

Sarason

Pierce

(1990) Social Support: An Interactional View. New York: John Wiley & Sons.

91.

Sato

Uono

Matsuura

Toichi

(2009) Misrecognition of facial expressions in delinquents. Child and Adolescent Psychiatry and Mental Health 3(1): 1–7.

92.

Schofield

Biggart

Ward

Larsson

(2015) Looked after children and offending: An exploration of risk, resilience and the role of social cognition. Children and Youth Services Review 51: 125–133.

93.

Skuse

Matthew

(2015) The trauma recovery model: Sequencing youth justice interventions for young people with complex needs. Prison Service Journal 220: 16–25.

94.

Spitzer

Siebel-Jürges

Barnow

Grabe

Freyberger

(2005) Alexithymia and interpersonal problems. Psychotherapy and Psychosomatics 74(4): 240–246.

95.

Suchy

Whittaker

Strassberg

Eastvold

(2009.) Facial and prosodic affect recognition among pedophilic and nonpedophilic criminal child molesters. Sexual Abuse: A Journal of Research and Treatment 21: 93–110.

96.

Taylor

Bagby

Parker

JDA

(1997) Disorders of Affect Regulation: Alexithymia in Medical and Psychiatric Illness. Cambridge: Cambridge University Press.

97.

UserVoice (2011) What’s your story? Young offenders’ insights into tackling youth crime and its causes. Available at: http://www.uservoice.org (accessed 24 December 2015).

98.

Van Goozen

SHM

Bowen

Main

(2013) Intervention directed at emotion recognition in delinquent adolescents. European Neuropsychopharmacology 23(Suppl. 2): S143.

99.

Van Goozen

SHM

Fairchild

Snoek

Harold

(2007) The evidence for a neurobiological model of childhood antisocial behavior. Psychological Bulletin 133: 149–182.

100.

Walsh

Scaife

Notley

Dodsworth

Schofield

(2011) Perception of need and barriers to access: The mental health needs of young people attending a Youth Offending Team in the UK. Health and Social Care in the Community 19(4): 420–428.

101.

Welsh Government/Youth Justice Board (2014) Children and young people first. Welsh Government/Youth Justice Board joint strategy to improve services for young people from Wales at risk of becoming involved in, or in, the youth justice system. Available at: https://www.gov.uk/government/publications/youth-justice-strategy-for-wales-children-and-young-people-first (accessed 9 June 2015).

102.

Wolf

Centifanti

LCM

(2014) Recognition of pain as another deficit in young males with high callous-unemotional traits. Child Psychiatry and Human Development 45(4): 422–432.

103.

Wright

Kamala

Field

(2011) Using bootstrap estimation and the plug-in principle for clinical psychology data. Journal of Experimental Psychopathology 2: 252–270.

104.

Zimet

Dahlem

Zimet

Farley

(1988) The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment 52: 30–41.

105.

Zimmermann

(2006) Delinquency in male adolescents: The role of alexithymia and family structure. Journal of Adolescence 29: 321–332.