ISSN 2232-2981 letnik/year 24, številka/number 2022/4 Journal of Criminal Justice and Security za teorijo in prakso zagotavljanja varnosti O reviji Varstvoslovje je znanstvena revija, ki spodbuja interdisciplinarno razpravo in izmenjavo ugotovitev s področja proučevanja varnosti. Prizadeva si osvetliti pravne, organizacijske, kriminobške, kriminalitetnopolitične, politološke, sociološke, psihološke in druge vidike varnostno relevantnih pojavov in konceptov. Revija prispeva h gbbljemu razumevanju vloge in delovanja skupnosti, organizacij in posameznikov, ki sodelujejo pri zagotavljanju varnosti. Primeri tematskih sklopov, ki jih revija pokriva, so: • teoretične osnove varnosti, • organizacija in upravljanje varnostnih mehanizmov, • zatiranje kriminalitete, • policijska dejavnost, • zasebnovarnostna dejavnost, • obveščevalno-varnostna dejavnost in • inprmacijska varnost About the Journal The Journal of Criminal Justice and Security is a scientific magazine fostering interdisciplinary discussion and exchange of findings in the field of safety and security studies. In its effort to shed light on legal, organisational, criminological, politobgical, sociological, psychological, and criminal-policy aspects oof security-relevant concepts and phenomena, it facilitates a deeper understanding of the voles and the functbning of society, organisatbns, and individuals cooperating in the provision of security. As such, the Journal of Criminal Justice and Security covers the folbwing thematic areas: theoretical foundations of safety and security, organisation and management of security mechanisms, combating criminality, policing, private security, intelligence and security, and inprmatbn security. Revija Varstvoslovje je odprtodostopna revija in izhaja v skladu s pogoji licence Creative Commons Attribution CC BY 4.0. Journal of Criminal Justice and Security is an open-access journal, published under the terms of the Creative Commons Attribution licence CC BY 4.0. Glavni in odgovorni urednik / Editor-in-Chef Urednik številk v slovenskem jeziku/Editor of Slovene Issues Urednika številk v angleškem jeziku / Editors of English Issues Tehnični urednici / Technical Editors Uredniški odbor /Editorial Board Marcelo Aebi, University of Laussane, Laussane, Switeerland Igor Areh, Univerza v Mariboru, Maribor, Slovenija Branko Ažman, Univerza v Mariboru, Maribor, Slovenija Hans-Juergen Albrecht, Max Planck Institute for Foreign and International Criminal Law, Freiburg, Germany Goran Basic, Linnaeus University, Vaxjo, Sweden Igor Bernik, Univerza v Mariboru, Maribor, Slovenija Tore Bjorgo, Norwegian Police University College, Oslo, Norway Gleb Bogush, Moscow State University - Lomonosov, Moscow, Russia Hans Boutellier, Verwey-Jonker Institute, Utrecht, The Netherlands Aleš Bučar-Ručman, Univerza v Mariboru, Maribor, Slovenija Jiri Burianek, Charles University, Prague, The Czech Republic Bojan Dobovšek, Univerza v Mariboru, Maribor, Slovenija loan Durnescu, University of Bucharest, Bucharest, Romania Katja Eman, Univerza v Mariboru, Maribor, Slovenija Chris Eskridge, University of Nebraska-Lincoln, Lincoln, NE, USA Benjamin Flander, Univerza v Mariboru, Maribor, Slovenija Loraine Gelsthorpe, University of Cambridge, Cambridge, UK Peter Grabosky, Australian National University, Canberra, Australia Beata Z. Gruszczynska, University of Warsaw, Warsaw, Poland Alistair Henry, University of Edinburgh, Edinburgh, UK Tim Hope, Keele University, Keele, UK Djordje Ignjatovič, University of Belgrade, Belgrade, Serbia Andrej Sotlar, Univerza v Mariboru, Maribor, Slovenija Andrej Sotlar, Univerza v Mariboru, Maribor, Slovenija Maja Modic, Univerza v Mariboru, Maribor, Slovenija Sanja Kutnjak Ivkovič, Michigan State University, East Lansing, MI, USA Ada Mendragič, Univerza v Mariboru, Slovenija Bernarda Tominc, Univerza v Mariboru, Slovenija Zoran Kanduč, Inštitut za kriminologijo, Ljubljana, Slovenija Klara Kerezsi, National Institute of Criminology, Budapest, Hungary Krysztof Krajewski, Jagelonian University, Krakow, Poland Gary LaFree, University of Maryland, College Park, MD, USA Rene Levy, GERN, Guyancourt, France Tomaš Loveček, University of Žilina, Žilina, Slovak Republic Alida V. Merlo, Indiana University of Pennsylvania, Indiana, PA, USA Mirjam Milharčič Hladnik, Znanstvenoraziskovalni center Slovenske akademije znanosti in umetnosti, Ljubljana, Slovenija Mahesh Nalla, Michigan State University, East Lansing, MI, USA Tinkara Pavšič Mrevlje, Univerza v Mariboru, Maribor, Slovenija Borislav Petrovič, University of Sarajevo, Sarajevo, B&H Paul Ponsaers, Ghent University, Ghent, Belgium Iztok Prezelj, Univerza v Ljubljani, Ljubljana, Slovenija Amedeo Recasens i Brunet, CESDIP, Paris, France Bojan Tičar, Univerza v Mariboru, Maribor, Slovenija Andromachi Tseloni, Nottingham Trent University, Nottingham, UK Gregor Urbas, Australian National University, Canberra, Australia Peter Weteels, Institute of Criminology, Hamburg, Germany Sabina Zgaga Markelj, Ustavno sodišče Republike Slovenije, Ljubljana, Slovenija Naslov uredništva / Editorial Address: Fakulteta za varnostne vede, Kotnikova ulica 8, 1000 Ljubljana, Slovenija Tel.: +386 (0)1 300 83 00, Fax: +386 (0)1 230 26 87, Internet: https://www.fvv.um.si/rV/ Vse članke, objavljene v reviji Varstvoslovje, recenzirata vsaj dva anonimna recenzenta. / All the articles submitted to the Journal of Criminal Justice and Security (Varstvoslovje) are peer reviewed by at least two anonymous peer reviewers. Revija varstvoslovje izhaja štirikrat letno (in sicer marca, junija, septembra in decembra)./ Journal of Criminal Justice and Security is published quaterly (in March, June, September, and December) Članki, objavljeni v reviji Varstvoslovje, so indeksirani v Criminal Justice Abstracts (EBSCO) in Criminal Justice Abstracts with Full Text (EBSCO). / Articles in the Journal of Criminal Justice and Security are abstracted and indexed in Criminal Justice Abstracts (EBSCO), Criminal Justice Abstracts with Full Text (EBSCO). Content Editorial ........................279 Uvodnik ........................280 ARTICLES Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj Identification of identical twins in police procedures ........................281 Nuša Crnkovič The Effectiveness of a Psycho-Educative Group Programme Regarding Relationships in the Treatment of Sexual Offenders: A Preliminary Study ........................298 Mirče Milenkov, Maja Modic »Stop and Search«: Slovenian Police Officers' Perspective ........................326 277 Editorial Dear readers! Present issue of the Journal of Criminal Justice and Security includes three papers, covering various criminal justice and security topics - the identification and treatment of twins in relation to criminality, psycho-educative group programme the treatment of sexual offenders, and police powers to stop and search. In the first paper, Miha Dvojmoc, Veronika Mrak Klavzar and Vanja Erculj review the treatment of identical twins in crimes and examine to what extent people notice the switch of identities, to what extent they can distinguish between twins, and what factors are related to the ability to distinguish them. Findings of their survey showed that age, but not gender or length of acquaintance, was related to the ability to distinguish between twins. The ability to distinguish between twins was positively correlated with the observation of confusion between them. The most common indicators for distinguishing between identical twins were found to be facial features, behavior, a distinctive voice, and recognizable accessories. Nusa Crnkovic presented the psycho-educative group therapy, developed at a high-security psychiatric hospital in the United Kingdom, in an attempt to provide a therapeutic input for sex offenders. The results of the study, conducted as a preliminary evaluation of the programme, suggest that there was a decrease in feelings of loneliness for both groups of patients that completed the URI programme, and a limited reduction of interpersonal difficulties. The findings suggest that the evaluated programme might have a limited effect on sex offenders. In their study on police powers to stop and search Mirce Milenkov and Maja Modic examine what attitudes police officers in Slovenia have towards the use of these powers. The findings of their study show that police officers are very well aware of the importance of their attitude when dealing with members of the public in police procedures. As police powers to stop and search are subject to strong criticism because of frequent cases of discriminatory use and links to racism and police violence, the authors suggest further research of this area in Slovenia. We are thankful to the authors for their contribution to the present issue by sharing the results of their interesting research endeavours. We wish you a pleasant and informative reading. Maja Modic, PhD Editor of English Issues 279 Uvodnik Spoštovani bralke in bralci! Tokratna izdaja revije Varstvoslovje vsebuje tri prispevke, ki naslavljajo raznolike varstvoslovne teme - identifikacijo in prepoznavo enojajčnih dvojčkov v povezavi s kriminaliteto, skupinsko psihoedukativno terapijo za obravnavo spolnih prestopnikov in t. i. stop & search policijska pooblastila. V prvem prispevku tokratne izdaje Miha Dvojmoč, Veronika Mrak Klavžar in Vanja Erčulj proučujejo obravnavo enojajčnih dvojčkov zaradi kaznivih dejanj in ugotavljajo, v kolikšni meri ljudje opazijo zamenjavo dvojčkov, v kolikšni meri so sposobni ločevati med njimi ter kateri dejavniki so povezani z zmožnostjo ločevanja. Ugotovitve njihove študije kažejo, da je z zmožnostjo ločevanja med dvojčki povezana njihova starost, ne pa tudi spol ali trajanje poznanstva. Najpogostejši indikatorji za razlikovanje med enojajčnimi dvojčki so obrazne poteze, način vedenja, značilen glas in prepoznavni dodatki. Nuša Crnkovic v svojem prispevku predstavi skupinsko psihoedukativno terapijo, ki so jo razvili v visoko varovani psihiatrični bolnišnici v Združenem kraljestvu, z namenom terapevtske obravnave spolnih prestopnikov. Rezultati študije, ki je bila izvedena kot preliminarna evalvacija programa, kažejo, da je prišlo do zmanjšanja občutka osamljenosti in omejenega zmanjšanja medosebnih težav pri obeh skupinah bolnikov, ki sta zaključili program. Na podlagi rezultatov se tako nakazuje omejena stopnja učinkovitosti obravnavanega programa za spolne prestopnike. V študiji o policijskih pooblastil stop & search sta Mirče Milenkov in Maja Modic ugotavljala, kakšen je odnos slovenskih policistov in policistk do uporabe teh pooblastil. Ugotovitve študije kažejo, da se policisti in policistke zelo dobro zavedajo pomena, ki ga ima njihov odnos pri obravnavi oseb v policijskih postopkih. Ker so pooblastila stop & search v Evropi in širše podvržena številnim kritikam, predvsem zaradi pogosto očitane in dokazane diskriminatorne uporabe ter posledično povezave z rasizmom in policijskim nasiljem, avtorja izpostavljata, da je smiselno to področje nadalje raziskovalno nasloviti tudi v Sloveniji. Avtorjem se zahvaljujemo, da so v svojih prispevkih z nami delili svoje zanimive raziskovalne ugotovitve. Želimo vam prijetno in poučno branje. Dr. Maja Modic Urednica številk v angleškem jeziku 280 Identification of identical twins Miha Dvojmoc, Veronika Mrak Klavzar, Vanja Erculj Purpose: The aim of the study was to review the treatment of identical twins in crimes and to determine to what extent people notice the switch of identities, to what extent they can distinguish between twins, and what factors are related to the ability to distinguish them. Design/Methods/Approach: We carried out an analysis of literature and a case review about identification of identical twins in police procedures and for the purpose of the article, an online survey was conducted among the general public. Findings: In line with existing research, the survey showed that age, but not gender or length of acquaintance, was related to the ability to distinguish between twins. The ability to distinguish between twins is positively correlated with the observation of confusion between them. Our survey of the public found that the most common indicators for distinguishing between identical twins were facial features, behavior, a distinctive voice, and recognizable accessories. Research Limitations/Implications: A review of the literature on the identification and treatment of twins in relation to criminality reveals several gaps in knowledge. More attention should thus be paid to the problem in the future, possibly to establish criteria for prosecution and to raise awareness that such deviant behavior can occur. In our research, we conducted survey among general public in the future maybe a survey of the experts should be conducted for more comprehensive view of the issue. Among the proposals for further research, it would be reasonable to consider the analysis of cases of identical twins in police practice. Originality/Value: In our own research, we examined the factors associated with identity swapping among identical twins. Keywords: identical twins, distinction, identity switch, fraud UDC: 343.525 in police procedures Justice and Security year 24 no. 4 pp.281-297 281 Identification of identical twins in police procedures Identifikacija enojajčnih dvojčkov v policijskih postopkih Namen: Namen prispevka je pregled obravnavanja enojajčnih dvojčkov pri kaznivih dejanjih in prikaz ugotovitev raziskave, v kolikšni meri ljudje opazijo zamenjavo identitete dvojčkov ter kateri dejavniki so povezani s sposobnostjo razlikovanja med njimi. Metode: Opravili smo analizo literature in pregled primerov o identifikaciji enojajčnih dvojčkov v policijskih postopkih, za namen prispevka pa smo izvedli tudi spletno anketo med širšo javnostjo. Ugotovitve: Ugotovili smo, da več kot polovica ljudi opaža zamenjavo identitete dvojčkov ter da je s sposobnostjo ločevanja med dvojčki povezana njihova starost, ne pa tudi trajanje poznanstva. Ravno sposobnost ločevanja med dvojčkoma je povezana z opažanjem zamenjave identitete, ne pa tudi spol dvojčkov. Raziskava je pokazala, da so najpogostejši identifikatorji razlikovanja med enojajčnimi dvojčki obrazne poteze, vedenje, značilen glas in drugi prepoznavni identifikatorji. Omejitve/uporabnost raziskave: Pregled literature o identifikaciji in obravnavi dvojčkov v povezavi s kaznivimi ravnanji razkriva več vrzeli v znanju. Problemu bi bilo zato treba v prihodnje posvetiti več pozornosti in morda vzpostaviti merila za pregon in ozaveščanje, da do tovrstnih deviantnih vedenj ne bi prišlo. V prihodnje bi bilo treba izvesti tudi raziskavo med strokovnjaki za bolj celovit pregled problematike. Izvirnost/pomembnost prispevka: V naši raziskavi smo preučili dejavnike, povezane z zamenjavo identitete med enojajčnimi dvojčki. Ključne besede: enojajčni dvojčki, ločevanje, zamenjava identitete, goljufanje UDK: 343.525 1 INTRODUCTION The problem of swapping and subsequent identification of twins involved in crimes could become increasingly important. According to experts (Vosnjak, 2018), the number of twins has almost doubled in recent years. The reasons for this are the increasing number of fertility treatment users and the increasing age of mothers. With identical twins, there is the possibility of fraud to gain advantages or avoid penalties. An example of this type of cheating is when one of the twins has a driving licence and the other does not, but they both use the same one. Identical twins are difficult to prosecute because of their identical DNA and appearance, and because of the law in Slovenia on the protection of human 282 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj rights and fundamental freedoms. Various methods such as facial recognition, fingerprint and palmprint matching, voice recognition, handwriting and iris recognition can be used to separate twins (Informacijski pooblaščenec [IPRS], 2013). The formation of identical and fraternal twins is influenced by various genetic and environmental factors. Identical twins occur in 3 to 4 cases per 1,000 births worldwide. Most cases of identical twins are not caused by genetic factors, although there are more cases of identical twins than usual in individual families (Spiro, 2015d). The issue of swapping and treatment of identical twins in criminal proceedings is poorly researched. There is little literature on the subject in Slovenia and in wider Europe. There are also gaps on this topic in the Criminal Code (»Kazenski zakonik (KZ-1-UPB2)«, 2021) or other laws. In our research, we therefore want to find out whether identical twins swap their identity to gain an advantage or escape sanctions, and whether this is related to the gender of the twins. We are interested in whether the ability to separate twins is related to the length of time the observer has known them, or to the age of the twins and the observation of the identity switching itself. We would like to identify the most common indicators that can be used to distinguish between identical twins. To this end, a survey among the general public and an analysis of media articles reporting on the confusion of twins in criminal cases were conducted. Biometrics is a science that deals with identifying people by their unique personal characteristics. The most common biometric distinguishing feature is the fingerprint, a snapshot of the papillary lines of the finger, and others are the individual characteristics of the eye (iris and retina), DNA, face, ears, and characteristic posture (IPRS, 2013), but individuals can also be distinguished on the basis of other physiological or physical characteristics and biological processes (these can, for example, also be based on other physiological and biological characteristics, such as the distance between parts of the face, papillary lines on hands and feet, shape of auricles, smell, body shape, height and weight, hair colour, shoulder width, leg shape) and behavioural characteristics (such as the way a person moves, walks, his/her tone of voice) (Pogačnik, 2013). Identical twins usually have very similar anatomy and appearance, but greater differences in appearance between them may also be the result of behavioural and epigenetic influences. Genetically, identical twins are identical - they have identical DNA because they are the result of the division of a single zygote (fertilised egg) at the beginning of pregnancy (after division, both fertilised eggs have identical DNA) (Juefei-Xu & Savvides, 2013). Biometric technology is known as a reliable and efficient identification and separation system, which is a product of technological development and represents a significant advance towards greater safety (Pogačnik, 2013). With our research, we wanted to take a first step towards exploring the above-mentioned issues in order to draw attention to possible shortcomings in the police process and contribute to their elimination in practice. The aim of the work is to highlight and investigate the frequency of the confusion of identical twins in 283 Identification of identical twins in police procedures various criminal proceedings, and the ability to distinguish identical twins in the general population. 2 DISTINCTION BETWEEN IDENTICAL TWINS Biometric verification is an increasingly common method of identifying individuals in the context of public safety, enterprise systems, and consumer electronics. It is used to verify the authenticity of certain characteristics of a person. Biometric data is data that relates to a specific or at least identifiable person. Fingerprints, for example, always belong to a specific person. Biometric data is stored in a central database, which often depends on local data collection methods. Gillis et al. (n. d.) explains that the main purpose of biometrics is security. Biometric security is a security mechanism that authenticates a person and secures access to a system or facility. It assesses a person's biological data or physical characteristics and is seen as the most seamless and strongest security technique for verifying a person's identity (Gillis et al., n. d.). It is primarily used in environments where higher critical physical security is required, or that are vulnerable to potential identity theft. Biometric security systems use various immutable characteristics of the human body. The physical characteristics of each person are stored in the biometric security system and can only be viewed by authorised personnel. When a person wants to access the system or enter a facility, the biometric scanner reads and evaluates the person's physical characteristics and matches them with the data stored about them. If this data matches, the system allows the person to access the system or enter the building or premises (Rouse, 2016). A very important feature of any biometric verification is its convenience, as users do not need a password or security token to use it. The components of any biometric device include a reader that verifies and records the presence of the biometric factor and software that converts the scanned biometric data into a standardised digital format and allows the comparison and matching of the captured data with the database where all biometric data is stored. A person is identifiable by means of an identification number or by means of one or more specific characteristics of their physiological and physical identity, the means of identification themselves being reasonably accessible not only to the controller but also to other persons. Biometric data can also be used to distinguish between identical twins. For example, identical twins do not have identical fingerprints (Spiro, 2015a). While the correlation of whorls, loops, and ridges between identical twins is high, there are differences where skin ridges meet, branch or end. Spiro (2015a) attributes the differences in fingerprints of identical twins to the interaction of different genes in the developmental environment of the uterus. Fingerprints have an important genetic component, but they also reflect the non-genetic environment of early pregnancy, an important time for tissue differentiation and organogenesis. In fact, the embryo begins to develop fingerprint patterns in the first weeks of gestation (Tao et al., 2012), and fingerprints are permanently configured before the twentieth week of gestation. 284 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj In addition to fingerprints, twins can also be distinguished by voice recognition. However, previous research suggests that the error rate in automatic voice recognition is higher in identical twins than the general population. Armour et al. (2014) stated that identical twins have similar vocal frequencies because the gross anatomy of the larynx and vocal cords is genetically determined. Similarities in speech production may have both genetic and developmental components due to proximity in childhood. Therefore, we would predict that identical twins would be more difficult to distinguish from each other in an identical environment than unrelated individuals (Van Gysel et al., 2001). Another of the biometric features used to verify identity is a person's handwriting. In addition to genetics and environment (Spiro, 2015c), several random factors influence a person's handwriting. Handedness, the tendency to write with the right or left hand, is genetically determined. Parents who are left-handed are more likely to have left-handed children. However, this is truer for siblings than identical twins. Many twins have opposite preferences for dominant use of a particular hand in writing. Srinivas et al. (2012) found that of all the ways to distinguish identical twins through biometric recognition and discrimination, facial recognition has received the most attention. Facial markers and the use of facial regions that develop with the twins' age are the most commonly used. This is because identical twins can look increasingly different as they age, since they are exposed to a more diverse environment (Spiro, 2015b). The results of a study by Mahalingam and Ricanek (2022) suggest that it is easier to distinguish identical twins by parts of the face rather than by the whole face. This is independent of gender (i.e., it works almost equally well for men and women), but is related to age. Research has shown that facial recognition becomes easier as twins get older. In the scientific literature, surprisingly little attention is paid to distinguishing identical twins with the naked eye. Given the results of biometric procedures for twin discrimination, we were interested in whether the distinguishing of twins is also possible with the naked eye, and whether it is related to the duration of acquaintance, age, and gender of the twins. 3 TWINS AND CRIMINAL TENDENCIES The topic is situated in various aspects of criminology, of which the biological perspective is the most important. The differences in the propensity to crime between identical and fraternal twins have already been researched in detail (Criminal Justice, n. d.). One of the first twin studies to genetically determine criminality was conducted in the 1920s by Johannes Lange (1929). He studied 30 pairs of identical twins. He found that in 10 out of 13 pairs of twins, both twins were known criminals. Subsequently, extensive studies were carried out to investigate whether at least one of each pair of twins had committed a crime. In 1974, Karl O. Christiansen (1974) studied the criminal behaviour of 3,586 pairs of twins born in Denmark between 1881 and 1910. He found that the probability of one twin being a criminal when the other twin was a criminal was 50% for identical twin pairs, but only 20% for pairs of fraternal twins. The correlation between the genetic 285 Identification of identical twins in police procedures proximity of biological relatedness and criminality was especially true for serious violent crimes and for longer criminal careers. These findings were confirmed by the results of a study of twin criminality in the 1980s and 1990s by Rowe (1983). This study showed that identical twins were more likely to be involved in criminal activity than fraternal twins. In addition, identical twins reported more criminal peers than fraternal twins (Rowe, 1983). The work of Rowe and his colleagues supports the genetic component of criminality, but also provides evidence of a social component. Although twin studies have provided some evidence for a genetic component in behaviour, it is difficult to separate the influence of genetics from that of social factors. This conclusion was also reached by Adler et al. (1998), who found delinquency in 32% of identical twins and in 12% of fraternal twins. Christiansen (1977) complemented the biological (genetic) view of deviant behaviour with an explanation from a psychological and social perspective. Mesko (2016) agreed, and explained that delinquency is not necessarily the result of a person's genetic predisposition, but can also simply be a disposition that develops in a certain environment. The influence of the family on later criminality (due to the influence of parents on child development) still belong in the realm of heredity in relation to criminality, but it is important to complement them with a psychological view of criminology, where, for example, social learning theory assumes that criminality is learned. Through observation, individuals can shape their own behaviour by imitation and reducing inhibitions about criminal behaviour, and if the latter is encouraged and rewarded, the effect is all the greater. Learning of behaviour through observation, direct exposure and various incentives (such as approval of such behaviour) occurs through the processes of identification and socialisation. In this process, the child identifies with their parents and adopts moral and other norms from them (Mesko, 2016). There is also a theoretical problem with the assumption that twins who grow up in the same parental home are exposed to the same treatment and social environment. Even scientists who study the link between criminal behaviour and genetics are cautious in their conclusions, arguing that these types of studies only show that similarities between twins influence behaviour. Whether these similarities are genetic or social, or a combination of both, and which of these explanations (broader social aspects based on similarity and socialisation, or individual-centred biological explanations involving genetic and psychological aspects) has a greater influence, is still controversial (Criminal Justice, n. d.; Newburn, 2007). Based on research findings, in our study we were interested in the extent to which crimes committed by identical twins were reported. To this end, we examined media coverage of the issue. Dealing with a crime committed by an identical twin is complicated by the strong resemblance and possible collusion between the twins (for reasons of protection or even intentional identity swapping). In this research we were therefore also interested in the extent to which the public observes identity swapping between twins to gain an advantage or avoid punishment, and whether the observed identity swapping between twins is in any way related to the gender of the twins. 286 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj 4 METHODS 4.1 Data collection To address the research objectives, an online survey was conducted. The survey was voluntary and anonymous. The sampling was non-random, more specifically snowball and purposive. The questionnaire consisted of two parts. The first set of questions was related to information about knowing and being able to distinguish between identical twins, and what the respondents knew about identity swapping in this group. When filling out the survey, the respondents were asked to refer to the identical twins they knew best. In the questionnaire, we also wanted to know if the respondents had ever been able to distinguish identical twins, and what were the most common characteristics they used to identify or distinguish them. The ability to distinguish between identical twins and the observation of identity swapping between them were measured by replies to statements using a five-point Likert scale of agreement. The ability to distinguish between twins was measured using seven (7) statements, namely the extent to which the respondents were able to distinguish between identical twins based on different styles of dress, different hairstyles, differences in facial features, differences in the rest of the body, differences in behaviour, handwriting, and voice. Perceptions of identity swapping or confusion between identical twins were measured using four statements and a five-point Likert scale of agreement. The following statements were included: I have observed identical twins swapping identities (for example, at school during tests, social events, in a car); I have thought (suspected) that identical twins have swapped identities; I have heard that identical twins have swapped identities; and It is known that identical twins have swapped identities. The last section of the questionnaire collected demographic information. 4.2 Sample description The sample comprised 223 respondents, of which women outnumbered men (there were 73.1% (163 respondents) women in the sample). Most (48.9%) of the respondents had a university degree (pre-Bologna and Bologna colleges and universities), while the fewest respondents had a PhD (1.3%). A total of 22.9% of respondents had completed secondary education, 13.9% of respondents had completed a pre-Bologna undergraduate degree or Bologna Master's degree, 9.4% had completed post-secondary education, and 3.6% had completed a master's degree. The average age of respondents was 34.7 years with a standard deviation of 13.7 years. The minimum age was 18 years, and the oldest respondent was 66 years old at the time of the survey. 287 Identification of identical twins in police procedures 4.3 Statistical analysis The validity of the measurements was checked by factor analysis, and the reliability by Cronbach's alpha. Factor analysis was performed using the principal axis method, applied separately for each set of statements on the ability to identify identical twins and on the observation of twin swapping. We calculated the composite variable as the average of the statements with high weights on each factor. Descriptive variables were represented by frequencies and proportions and numerical variables by the arithmetic mean and standard deviation. Chi-square test and multiple linear regression were used to test the hypotheses. All hypotheses were tested with a significance level of a = 0.05. A content analysis of news reports on the topic of fraud with identical twins was also conducted Newspaper and website articles on the topic of fraud with identical twins were reviewed. The articles were searched for in the Slovenian database of Google with the keywords »enojajčni dvojčki, goljufanje, ločevanje, zamenjava identitet«. We searched for data on such fraud abroad using the keywords identical twins, fraud, separation, change of identity. All the search results were examined and only those reporting the criminal act of identical twins and accompanying police procedures were included in the analysis. The analysis of the articles included information about the country and year in which the offence of the identical twins was committed, gender of the identical twins and the provability of the offence, and summarised the content of each case of the offence. 5 RESULTS 5.1 Recognition of identical twins 5.1.1 Validity and reliability of measurement Factor analysis revealed that two factors could be identified: the ability to identify identical twins, measured by seven (7) statements on the Likert scale, and the observation of identity confusion, measured by four (4) statements on the Likert scale. Due to the small sample size, factor analysis was conducted separately for each set of statements. These statements were combined into two factors or composite variables that were averaged across the statements with high weights on each factor and used in the rest of the study. The data on the first set of statements about the ability to distinguish between twins are suitable for factor analysis (KMO = 0.86; x2(21) = 937.2; p < 0.001). All statements have a high weight (0.69-0.82) on a single factor explaining 63.6% of the variability of the measured variables. Reliability, as measured by Cronbach's alpha for the discrimination factor, is 0.90. The KMO measure and Bartlett's test for sphericity indicate that the data related to the observation of identity confusion are suitable for factor analysis (KMO = 0.75, x2(3) = 620.7; p < 0.001). Two statements are perfectly correlated (I suspect swapping and I notice swapping), so we include only one, I suspect confusion, in the analysis. All statements have a high weight on a factor (0.88- 288 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj 0.94). The reliability of the measurement of the factor »I notice swapping« was 0.93. 5.1.2 Identical twins and how to distinguish between them Respondents were mostly friends with identical twins (43%), followed by acquaintances (23.1%), and least likely »other« (7.3%) and (former) colleagues (2.7% of respondents). Under the option »other«, respondents named ex-friend, I am an identical twin, partner, neighbour, ex-boyfriend, teacher, child of an acquaintance, know from kindergarten, brother, father, and brother of a friend (two times each). Table 1 shows that the average age of acquaintances of identical twins is 11.81 years, with a standard deviation of 9.42 years, and the average age of twins is 29.71 years, with a standard deviation of 13.69 years. Almost two thirds of the respondents know female identical twins (data not shown). Average value Standard deviation Years of acquaintance 11.81 9.42 Age of twin 29.71 13.69 Facial features Behaviour Distinctive accessories m Voice o 4-" m Hairstyle 7 o T3 Body posture 7 Way of walking ^^^m 5.1 Do not distinguish ^^m 3.8 Handwriting H 1.8 Other ™ l.S Percentage shares Table 1: Years of acquaintance and age of twins (n = 221) The survey found that the most common indicators for recognising identical twins are facial features, behaviour and voice, and distinctive accessories (Graph 1). Other possible identifiers include weight, tattoos, stature, different body sizes when standing together, different points on the hands, knowledge, scar, small mole on the face, and brow ridge injury or scar over the eye, and birthmarks. The average agreement on the ability to distinguish between twins (factor or composite variable) was 2.93 and the standard deviation was 0.96. The average on a five-point scale shows that respondents were quite uncertain about their ability to distinguish between identical twins. Figure 1: Identification indicators for identical twins (n = 221) 289 Identification of identical twins in police procedures 5.1.3 Factors associated with the identification of identical twins Multiple linear regression was used to determine the relationship between length of acquaintance with identical twins, age of identical twins and ability to identify (distinguish) identical twins. The results of the regression analysis are shown in Table 2. The years of acquaintance are not statistically significantly related to the ability to distinguish between twins (p = 0.32). However, controlling for the length of acquaintance with identical twins, we find that there is a statistically significant relationship between the age of the twins and the ability to distinguish between them (p < 0.001). The positive value of the regression coefficient indicates that the older the twins, the greater their ability to distinguish between them. Table 2: Correlation between the length of acquaintance, the age of the identical twins and the ability to distinguish between them (n = 221) Regression coefficient Std. error P (Constant) 0.87 0.14 <0.001 Years of acquaintance 0.01 0.01 0.32 Age of twin 0.03 0.01 <0.001 R2 = 0.17 5.1.4 Observation of swapping of identical twins and factors related to the observation of swapping We were interested in the proportion of people who notice identity swapping between identical twins. We wanted to check whether more than half of the people noticed such a case with identical twins. The analysis included the statement »I have noticed identity swapping between identical twins«, where respondents were classified as not noticing if they chose the answer »I do not agree at all« and as noticing if they chose other answers on the agreement scale. We found that 73.5% of the respondents had noticed the swapping of identical twins. This proportion is statistically significantly different from 50% (x2 (1) = 48.4; p < 0.001). Observation of identity swapping (composite variable as the average of statements measuring observation of the swap) is not statistically significantly (p = 0.66) associated with twin gender (Table 3), but is statistically significantly associated with the ability to distinguish between twins (p < 0.001). Individuals who rated themselves as having a better ability to distinguish between twins also observed identity swapping to a greater extent. Table 3: Relationship among gender, the ability to distinguish identical twins, and observation of identity swapping (n = 221) Regression coefficient Std. error P (Constant) 0.86 0.17 <0.001 Male twins 0.05 0.10 0.66 Ability to distinguish 0.32 0.07 <0.001 R2 = 0.12 290 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj 5.2 Content analysis of the cases of fraud and dealing with criminal offences of identical twins Table 4 shows some of the characteristics of the articles that reported on identity swapping between identical twins. Two of the articles were from Slovenia, and one each from Bosnia & Herzegovina, China, Sweden, and France. All articles discussed male twins, and only in two cases were the investigating authorities able to prove their guilt. Country Year of criminal activity (CA) Form of CA Gender of identical twins Age of identical twins CA has been proved / has not been proved Slovenia 2011 Theft Male Unknown Not proved Slovenia 2005 Causing serious injury Male Unknown Not proved China (year unknown, fraud lasted for 20 years) Identification fraud (both twins used one driving licence) Male Unknown Proved Sweden 2012/2013 Suspicion of murder Male Unknown Not proved Bosnia & Herzegovina 2019 Suspicion of murder Male 67 Not proved France 2012/2013 Three rapes and three attempted rapes Male Unknown Proved (one of the identical twins has confessed to the crimes, but it is not certain if he is the right offender, due to both men having the same DNA) Table 4: Analysis of cases of fraud and the treatment of identical twin crimes in Slovenia and abroad (newspaper articles) All these cases involved only the above-mentioned offences and not identity fraud. In Slovenia, two cases of offences involving one or both twins were reported. In the case of the trial of identical twins E.E: and A. E. they were charged with burglary and theft in a commercial building. The burglars were intercepted by police and then fled, leaving behind a car with several burglary tools. Forensic investigators found a biological trace in the perpetrator's car belonging to one of the twin brothers. The lawyers for the twin brothers claimed that criminal proceedings can only be initiated and carried out against a specific person, whereas in the brothers' case it is not known and cannot be established which of them is supposed to have left the biological traces. This is »an extremely unusual case arising from the fact that Alen and Edin Eljezi are identical twins who are believed to have identical STR profiles, i.e. repeating DNA strands« (Predanic, 2016). This case shows how complicated criminal proceedings can be when identical twins are in the dock. In Slovenia, we are also familiar with the case of the trial of an identical twin, the security guard B. L., who allegedly inflicted serious bodily harm to the head of a nightclub guest. Criminal proceedings have been initiated against Lesjak, who denies his guilt and claims that he was at another location on the night in questionand was not working at the club. On the day of the trial, 291 Identification of identical twins in police procedures Boštjan's identical twin brother appeared in court instead of him and made a false statement. Forensic identification during the investigation revealed that the victim had identified Boštjan Lesjak as his attacker. The investigators did not find any DNA on the perpetrator that would have enabled a reliable identification in a conventional DNA analysis. However, experts consider DNA analysis of identical twins to be a grey area, as they have the same DNA profile. However, since the accused had an identical twin who looked very similar to him, it would also be difficult for the prosecution to gather enough evidence for a conviction, which in fact they failed to achieve in this case (Furlan Rus, 2014). Examining data from around the world, we came across the case of the Yan identical twins in China, where the brothers held the same Class B driving licence for 20 years (A. P., 2019). In France, we came across the case of the Gomis trial against identical twin brothers who were convicted of several offences. They were charged with three counts of rape and three counts of attempted rape. Forensic experts did their best, but because the brothers had almost identical DNA the French police took them both into custody. The brothers were very close, lived in the same flat, wore the same clothes, used the same car and phone, and shared a Facebook profile. The rape victims were able to identify the perpetrator in court, but they were unable to identify the perpetrator or distinguish which of the brothers was the actual perpetrator. One witness identified Yoan as the perpetrator based on his type of speech. Meanwhile Elvin was released, as there was not enough evidence to convict him. After ten months in prison, Yoan broke down and regretted being the 'evil twin' and thus saving his brother (G. G., 2017). In Sweden (Furlan Rus, 2019), there was a trial of identical twins who were in custody on suspicion of murder. Since they shared DNA it was not possible to identify them from the DNA traces found at the crime scene on a gun cartridge and on cigarette butts found in a jar in the murdered man's car. Since it was never discovered whose DNA was found, the pair of twins were never convicted. Moreover, if it was only one of the twins then clearly they cannot both be tried, because then one of them would be wrongly convicted. Elsewhere there are further cases of identical twins with identical DNA suspected of serious crimes. While researching newspaper articles, we also came across the case of the Stankovic identical twins, who are on trial in Bosnia & Herzegovina for the murder of a dentist by shooting him in the chest. A DNA trace was found at the crime scene that is believed to belong to the killer. The trace matches the DNA profile of the arrested brothers. Since they are identical twins, and have the same DNA, it remained to be clarified which of them committed the crime. Dejan is suspected of being the murderer, but the investigation is still ongoing. Later in the court proceedings, the authorities could not prove the guilt of either one or the other of the identical twins (Milinkovic & Subota, 2015). 6 DISCUSSION The cases of trials of identical twins show that there are repeated reports in the media about the confusion of male identical twins, and that both investigators and prosecutors face obstacles in establishing the identity of the perpetrator in 292 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj such cases. The problem here is because identical twins have identical DNA. However, there are certain techniques, such as biometrics, that make it possible to establish identity based on other characteristics. IPRS (2013) states that the most used biometric characteristics are fingerprints, DNA, iris and retina, ears, face and characteristic body posture. The characteristics of a person include physiological body features and behavioural traits such as movement, gait, eyes, and voice (Pogačnik, 2013). In the case of identical twins, it is impossible to determine the true identity of a person just by looking at a document, since they will have the same photo, same date of birth, and often the same residential address. The main challenge in face recognition for identical twins is to find an approach for feature extraction and to formulate the separation of identical twins in space. It is important to find out which facial components - such as the eye area, nose, or mouth - are the best distinguishing features. A study (Mahalingam & Ricanek, 2022) has shown that recognition performance based on eye area almost equals or even surpasses recognition performance based on the whole face. Sharing the same DNA profile also makes it difficult for investigators to biometrically identify identical twins, as they usually only secure DNA traces at the crime scene. When investigating a single crime where the suspects could be identical twins, additional evidence and information must be gathered and the alibis of both twins must be verified. Only through comprehensive investigations is it possible to produce evidence that allows conclusions to be drawn as to which of the twins committed the crime. When investigating a case where it is necessary to distinguish between two adult twins and to know who is responsible for the crime, the police must look for their fingerprints along with any scars, identifying marks or tattoos, in addition to observing the circumstances of the crime where both or at least one of the identical twins was present (Souza, n. d.). Our survey of the public found that the most common indicators for distinguishing between identical twins were facial features, behaviour, a distinctive voice, and recognisable accessories. Other distinguishing features included differences in weight, which are usually evident in a difference in stature, and distinctions based on tattoos, different marks on the hands and face, and other signs. It is difficult to distinguish identical twins with the naked eye, as we have found in our own research. Our research has shown that the ability to distinguish between twins is related to the age of the twins, as in some other studies (Mahalingam & Ricanek, 2022), but not to the length of acquaintance. This is in contrast to some research suggesting that greater differences between twins occur in adulthood because they are exposed to a more diverse environment, as well as differences in twins' lifestyles, such as exercise, diet, smoking, alcohol consumption, employment, body size, presence of a spouse, and children (Spiro, 2015b). Siblings often start out with very similar personality traits and intelligence, but slowly drift apart over time. E. Turkheimer, , believes this is because small differences that occur early in life - perhaps due to a chance event or genetic change - somehow increase as we grow up (Hamilton, 2013). As an example, he cites identical twin girls, one of whom is slightly more extroverted than her sister due to a random event, which leads to her meeting different people, making 293 Identification of identical twins in police procedures more extroverted friends and getting a different kind of partner to her sibling. The result is a feedback loop that continuously reinforces this personality trait, and possibly even generates new brain cells. Indeed, Turkheimer conducted a study on mice that showed how small changes in behaviour can lead to larger changes and, as a result, even change the brain (Hamilton, 2013). In our own research, we examined the factors associated with identity swapping among identical twins. This was noted by more than half of the respondents in the survey, and the result could be generalised to the wider population. Identity confusion was not statistically significantly related to the gender of the twins, but it was related to the ability to tell them apart. The work of Rowe et al. (1983) supports the genetic component of criminality, but also provides evidence of a social component (social factors) that is difficult to separate. Another study examining criminality among identical twin pairs of the same gender, found that in most cases both twins were known criminals (Lange, 1929). Subsequently, extensive studies were carried out to investigate whether at least one of each pair of twins had committed a crime. It was found that there was a 50% chance that one of the identical twins would get involved in criminal activity, when the other twin was already a criminal, and only 20% for fraternal twin pairs (Christiansen, 1974). The correlation between the genetic proximity of biological relatedness and criminality was especially true for serious violent crimes and for longer criminal careers. A review of the literature on the identification and treatment of twins in relation to criminality reveals several gaps in knowledge. More attention should thus be paid to the problem in the future, possibly to establish criteria for prosecution and to raise awareness that such deviant behaviour can occur. Women are now choosing to have children later and later in life, leading to increased use of artificial insemination and higher numbers of identical twins (Spiro, 2015d). Identical twins often exploit their similarity for various purposes, including avoiding punishment. A police officer establishes a person's identity with authority and under certain conditions, and to exercise the power of identification, the police officer must assess whether any grounds for suspicion exist. Based on our research, we conclude that swapping of identity by identical twins in relation to police procedures should be regulated by law. For example, a note could be included in a driving licence stating that the person has an identical twin, as in the case of needing to wearing glasses, which would make it mandatory to identify the person by taking a fingerprint. It would be possible to use an application that scans the fingerprint to identify the person's data and verify the match with the ID document. This would help to identify and distinguish identical twins in police procedures, especially for traffic police better and more easily. This would prevent police officers from being held criminally responsible for improperly conducted procedures resulting from fraud by identical twins. In the future, it would be useful to extend this research to other areas of police investigations, especially criminal investigations where only DNA traces are available, if that. The denial of both identical twins can leave the case open, so the very consideration of this issue would help ensure that proper DNA and other identification would bring the real perpetrators of such crimes to justice. 294 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj REFERENCES A. P. (9. 4. 2019). Dvojčka imela le eno vozniško dovoljenje [Twins only had one driver's license]. Slovenske novice. https://www.slovenskenovice.si/kronika/ na-tujem/dvojcka-imela-le-eno-voznisko/#!?logout Adler, F., Mueller, G. O. W., & Williams, L. S. (1998). Criminology. McGraw-Hill. Armour, J. A., Davison, A., & McManus I. C. (2014). Genome-wide association study of handedness excludes simple genetic models. Heredity, 112(3), 21-25. https://doi.org/10.1038/hdy.2013.93 Christiansen, K. O. (1974). Seriousness of criminality and concordance among Danish twins. In R. Hood (Ed.), Crime, criminology, and public policy (pp. 6377). Heinemann. Christiansen, K. O. (1977). A preliminary study of criminality among twins. In S. Mednick & K. O. Christiansen (Ed.), Biological causes of human behaviour. Gardner Press. Criminal Justice. (n. d.). Biological theories of crime. https://criminal-justice. iresearchnet.com/criminology/theories/biological-theories-of-crime/15/ Furlan Rus, M. (30. 1. 2014). Zmeda zaradi dvojčkov [Confusion over twins]. Dnevnik. https://www.dnevnik.si/1042621411 Furlan Rus, M. (29. 8. 2019). Enojajčna dvojčka molčita, priča na Švedskem tudi [Identical twins are remaining silent, as well as witness in Sweden]. Dnevnik. https://www.dnevnik.si/1042896451 G. G. (9. 8. 2017). Policija je v zagati, ko sta posilstva osumljena enojajčna dvojčka [The police are in quandary when two identical twins are suspected of rape]. Svet24. https://novice.svet24.si/clanek/novice/crna-kronika/55eea9ad2a33e/ policija-je-v-zagati-ko-sta-posilstva-osumljena-enojajcna-dvojcka Gillis, A. S., Loshin, P, & Cobb, M. (n. d.). Biometrics. Techtarget. https:// searchsecurity.techtarget.com/definition/biometrics Hamilton, J. (9. 5. 2013). How can identical twins turn out so different?. NPR. https://www.npr.org/sections/health-shots/2013/05/14/182633402/how-can-identical-twins-turn-out-so-different?t=1645094438143 Informacijski pooblaščenec [IPRS]. (2013). Biometrija [Biometrics]. https:// www.ip-rs.si/varstvo-osebnih-podatkov/obveznosti-upravljavcev/prijava-biometrijskih-ukrepov/ Juefei-Xu, F., & Savvides, M. (2013). An augmented linear discriminant analysis approach for identifying identical twins with the aid of facial asymmetry features. In IEEE Conference on computer vision and pattern recognition workshops (pp. 56-63). CyLab Biometrics Center, Department of Electrical and Computer Engineering Carnegie Mellon University. https://www.cv-foundation. org/openaccess/content cvpr workshops 2013/W02/html/Juefei-Xu An Augmented Linear 2013 CVPR paper.html Kazenski zakonik [Criminal Code] (KZ-1-UPB2). (2012, 2015, 2016, 2017, 2020, 2021). Uradni list RS, (50/12, 54/15, 6/16, 38/16, 27/17, 23/20, 91/20, 95/21, 186/21). Lange, J. (1929). The importance of twin pathology for psychiatry. Allgemeine Zeitschrift für Psychiatrie und Psychisch- Gerichtliche, 90, 122-142. Mahalingam, G., & Ricanek, K. Jr. (2022). Investigating the effects of gender and age 295 Identification of identical twins in police procedures group based differences in identical twins. Institute of Interdisciplinary Science in Identityin Identity Sciences (Face Aging Group). University of North Carolina Wilmington. Meško, G. (2016). Kriminologija [Criminology]. Fakulteta za varnostne vede. Milinkovic, S., & Subota, T. M. (29. 3. 2015). DNK misterija: Ubica se krije iza brata blizanca [DNA mystery: The killer hides behind the twin brither]. Blic. https:// www.blic.rs/vesti/hronika/dnk-misterija-ubica-se-krije-iza-brata-blizanca/ sdlewhs Newburn, T. (2007). Criminology. Willan Publishing. Pogačnik, N. (2013). Biometrična tehnologija kot element varnosti v letalskem prometu [Biometric technology as an element of security in air transport] [Diplomsko delo]. B&B izobraževanje in usposabljanje, višja strokovna šola. Predanič, J. (5. 12. 2016). Ena biološka sled, dva obtožena [One biological trace, two accused]. Delo. https://old.delo.si/novice/kronika/ena-bioloska-sled-dva-obtozena.html Rouse, M. (11. 11. 2016). Biometric security. Techopedia. https://www.techopedia. com/definition/6203/biometric-security Rowe, C. D. (1983). A biometrical analysis of perceptions of family environment: A study of twin and singleton sibling kinships. Child Development, 54(2), 416— 423. Souza, D. (n. d.). How to tell identical twins apart?. Invivo. http://www.invivo. fiocruz.br/cgi/cgilua.exe/sys/start.htm?UserActiveTemplate=english&info-id=1421&sid=43 Spiro, R. (1. 10. 2015a). Do identical twins have identical fingerprints?. Washington State Twin Registry. https://wstwinregistry.org/2015/10/01/do-identical-twins-have-identical-fingerprints/ Spiro, R. (1. 10. 2015b). Can identical twins look different?. Washington State Twin Registry. https://wstwinregistry.org/2015/10/01/can-identical-twins-look-different/ Spiro, R. (1. 10. 2015c). Do identical twins always have the same hand preference?. Washington State Twin Registry. https://wstwinregistry.org/2015/10/01/do-identical-twins-always-have-the-same-hand-preference/ Spiro, R. (1. 10. 2015d). How often do twin birth occour?. Washington State Twin Registry. https://wstwinregistry.org/2015/10/01/how-often-do-twin-births-occur/ Srinivas, N., Aggarwal, G., Flynn, P. J., & Vorder Bruegge, R. W. (2012). Analysis of facial marks to distinguish between identical twins. IEEE Transactions on Information Forensics and Security, 7(5), 1536-1550. Tao, X. Chen, X., Yang, X., & Tian, J. (2012). Fingerprint recognition with identical twin fingerprints. PLoS ONE, 7(4), e35704. https://doi.org/10.1371/journal. pone.0035704 Van Gysel, W. D., Vercammen, J., & Debruyne, F. (2001). Voice similarity in identical twins. Acta Oto-rhino-laryngol Belgium, 55(1), 49-55. Vošnjak, A. (13. 4. 2018). Z dvojčki skoraj podvojeni trebuščki, slabosti in zapleti [Whit twins, almost double the tummies, weaknesses and complications]. Dnevnik. https://www.dnevnik.si/1042818299 296 Miha Dvojmoč, Veronika Mrak Klavžar, Vanja Erčulj About the Authors: Miha Dvojmoč, PhD, Associate Professor, Faculty of Criminal Justice and Security, University of Maribor, Slovenia. E-mail: miha.dvojmoc@um.si Veronika Mrak Klavžar, M.A., E-mail: veronika.mrak@zvil.si Vanja Erčulj, PhD, Assistant Professor, Faculty of Criminal Justice and Security, University of Maribor, Slovenia. E-mail: vanja.erculj@um.si 297 TZri'E The Effectiveness of a PsychoJustice and Security ^ To2i Educative Group Programme finn n - pp.298-325 Regarding Relationships in the Treatment of Sexual Offenders: A Preliminary Study Nusa Crnkovic Purpose: At a high-security psychiatric hospital in the UK, the Understanding Intimacy and Relationships (URI) psycho-educative group therapy was developed in an attempt to provide a therapeutic input for interpersonal difficulties displayed by patients with sex offending history. This study aimed to conduct a preliminary evaluation of the URI group effectiveness for sex offenders. Design/Methods/Approach: A longitudinal study with three groups of participants - sex offenders (n = 9) and non-sex offenders (n = 9) that completed the URI group, and the control group (n = 10), which were assessed at two time points was conducted at a high-security psychiatric hospital. Each group filled out three questionnaires at both timepoints - IIP-C, UCLA Loneliness Scale, and ECR-R. Findings: The results suggest a decrease in feelings of loneliness for both groups of patients that completed the URI programme, and a limited reduction of interpersonal difficulties. Results of within-subject changes regarding interpersonal difficulties among patients in URI group did not statistically significantly differ from the result of the control group. Research Limitations/Implications: The most evident limitation of the study is a very small sample size and lack of objective measurement of patients' difficulties in interpersonal relations. Practical Implications: The findings suggest that the URI programme might have a limited effect on sex offenders due to the unsuccessful implementation of the Risk, Need, Responsivity Model, although several study limitations were present potentially affecting the outcome. 298 Nuša Crnkovič Originality/Value: Longitudinal evaluation effectiveness of a clinical intervention for patients with sex offending history at a high-secure psychiatric hospital. Keywords: sexual offenders, attachment style, loneliness, psycho-educative group therapy UDC: 159.964.227:343.541-055.1 Učinkovitost psihoedukativnega skupinskega programa za medosebne odnose pri zdravljenju spolnih prestopnikov: preliminarna študija Namen prispevka: V visoko varovani psihiatrični bolnišnici v Veliki Britaniji (VB) so razvili skupinsko psihoedukativno terapijo Razumevanje intimnosti in odnosov (angl. Understanding Intimacy and Relationships (URI)) z namenom znižanja izraženosti težav v medosebnih odnosih, izraženih pri pacientih z zgodovino spolnega prestopništva. Namen raziskave je bilo izvesti preliminarno oceno učinkovitosti URI za spolne prestopnike. Metode: V visoko varovani psihiatrični bolnišnici v VB je bila izvedena longitudinalna študija, ki je vključevala tri skupine - udeleženci, ki so zaključili URI z zgodovino spolnega prestopništva (n = 9) in brez (n = 9) ter kontrolno skupino pacientov (n = 10), ki so bili ocenjeni na dveh časovnih točkah. Vsaka skupina udeležencev je izpolnila tri vprašalnike - IIP-C, lestvica osamljenosti UCLA in ECR-R. Ugotovitve: Rezultati nakazujejo na zmanjšanje občutka osamljenosti pri obeh skupinah, ki sta zaključili program URI ter omejeno zmanjšanje medosebnih težav. Rezultati niso pokazali statistično pomembnih sprememb na področju medosebnih težav pri posameznikih od prve točke testiranja do druge točke testiranja v kateri koli od treh skupin udeležencev. Omejitve/uporabnost raziskave: Ključna omejitev je majhen vzorec in pomanjkanje objektivnega merskega instrumenta pacientovih težav na področju medosebnih odnosov. Praktična uporabnost: Na podlagi rezultatov se tako nakazuje omejena stopnja učinkovitosti programa URI za spolne prestopnike zaradi neuspešne implementacije modela Tveganje, potrebe, odzivnost, vendar so predstavljene tudi omejitve študije, ki so lahko vplivale na rezultate. 299 The Effectiveness of a Psycho-Educative Group Programme ... Izvirnost/pomembnost prispevka: Longitudinalna ocena učinkovitosti klinične intervencije za paciente z zgodovino spolnega prestopništva v visoko varovani psihiatrični bolnišnici. Ključne besede: spolni prestopniki, oblika navezanosti, osamljenost, psihoedukativna skupinska terapija UDK: 159.964.227:343.541-055.1 1 INTRODUCTION Intimacy is one of the fundamental human needs that we require for general functioning (Popovic, 2005). Although there is no conclusive definition of intimacy, it can be broadly considered as a relational experience characterised by mutual exchange and proximity, the degree of which is contingent on the individual's perception of the experiences in a relationship (Marshall, 1989; Wynne & Wynne, 1986). The capacity to experience intimacy and form intimate relationships in adulthood is dependent on the attachment style, which first develops with a primary caregiver in childhood (Bowlby, 1973). More specifically, how a person interacts with his or her primary caregiver in infancy and childhood may influence the future personality organisation of that individual. With various experiences in different stages of life, the attachment style continues to develop and modify. It reflects one's manner of formation and engagement in interpersonal relationships (Mikulincer & Shaver, 2007). The attachments styles are broadly classified as secure and insecure. If the child experiences responsive and consistent care as well as supportive relationships, he or she will be more likely to develop internalised relational strategies, which will enable him or her to develop a coherent sense of self and maintain interpersonal relationships (Sroufe et al., 2009). On the other hand, if the child is exposed to malignant early caregiving, the child's ability to organise and regulate internal affects and cognitions may be distorted on several levels. Consequently, the exposure to interpersonal experiences characterised predominantly by negative emotions may impair the child's organisation and integration of self and disrupt the ability to form and/ or maintain relationships in future (Carlson & Ruiz, 2016; Sroufe et al., 2009). Research results suggest that there are various forms of insecure attachment styles, depending on the one's experience with their primary care giver. More specifically, anxious or insecure-ambivalent attachment style, avoidant attachment style and disorganised attachment styles (Ainsworth & Bell, 1981; Ainsworth et al., 2015; Groh et al., 2017; Main & Solomon, 1990). The anxious attachment style, also referred to as insecure-ambivalent style, is characterised by simultaneous dependant and rejecting behavioural response pattern in interpersonal relationship - e.g. despite the desire to form a close intimate relationship one will be vary to do so. Individuals with the avoidant attachment style will exhibit high level of physical and emotional independence as interpersonal relationships bring them discomfort (Ainsworth & Bell, 1981; Ainsworth et al., 2015). Lastly, the disorganised attachment style reflects lack of coherent or consistent behaviour 300 Nusa Crnkovic in interpersonal relationship and is often seen in those who were maltreated or abused by their caregivers (Main & Solomon, 1990). An inability to form and/or sustain intimate relationships as an adult is highly prevalent among individuals with a history of childhood trauma and is linked to a higher risk of developing psychiatric disorders in the future (Bowlby, 1973; Carlson & Ruiz, 2016; Carr et al., 2018; Cuoco et al., 2021; Maniglio, 2011; Marshall, 1989; Ozcan et al., 2016; Waldinger et al., 2006). Moreover, Marshall (1989) proposed in his theory that an insecure attachment style can also be understood as the absence of intimacy and the presence of loneliness. Extant research shows a high prevalence of childhood neglect, physical and sexual abuse, experiences of intense feelings of loneliness, lack of social contact and long-term intimate relationships among offenders with mental disorders (Craissati, 2009; Levenson et al., 2015; Meloy & Gothard, 1995; Rice & Harris, 1997). In line with this, there is a substantial body of evidence supporting the relationship between an insecure attachment style, mental disorders, and offending behaviour (Armstrong & Mellor, 2016; Grady et al., 2018; Levinson & Fonagy, 2004; Ogilvie et al., 2014; Smallbone & Dadds, 1998; Van IJzendoorn et al., 1997; Ward et al., 1996). 1.1 Attachment style, intimacy and loneliness in sex offenders The above findings appear relevant for sex offenders in particular as there is a high prevalence of childhood physical and sexual abuse among this group of offenders, thus raising the likelihood of suffering consequences such as insecure attachment and related feelings of loneliness (Jespersen et al., 2006). Studies have found a high prevalence of insecure attachment amongst sex offenders (e.g. in one study 75% of sex offenders reported insecure attachment style in adulthood), more fearfulness of intimacy and rejection compared to other groups of offenders (Bumby & Hansesn, 1997; McCormack et al., 2002; Ward et al., 1997). They tend to engage less in practices that stimulate intimacy, which is associated with low self-esteem, shame, loneliness and use of sexual activities as means of coping with stressors - all of which are linked to sexual offending (Bumby & Hansen, 1997; Cortoni & Marshall, 2001; Marshall et al., 2009). However, despite certain similarities in their interpersonal difficulties among sex offenders, researchers report that different types of sex offenders have been found not to share similar early interpersonal experiences and thus, have distinct attachment styles. Rapists compared to child molesters had fewer boundaries set by their caregivers, were subjected to more physical abuse and felt less safe. Child molesters, on the other hand, are two times more likely to report the presence of sexual abuse in childhood than rapists (McCormack et al., 2002; Seghorn et al., 1987). Furthermore, rapists' sexual aggressive behaviour does not necessarily reflect their struggle to form social bonds, but rather their preference to avoid intimacy with close-ones and avoidant attachment style. In contrast, child molesters' intimacy issues may be explained by their fear of intimacy, social anxiety and anxious attachment style (Martin & Tardif, 2014; Ward et al., 1996). Sexual recidivism has been associated with the presence of problematic ways of 301 The Effectiveness of a Psycho-Educative Group Programme ... relating to other people and specific patterns adopted when engaging in intimate relationships (Thornton, 2002). Insecure attachment, intimacy deficits, and loneliness as significant aspects of sexual offending behaviour were first proposed by Marshall (1989, 1993) and then further elaborated by Marshall and Marshall (2000). They theorised that insecure attachment has a negative effect on a child's development of self-esteem and flexible social skills resulting in an inability to engage in satisfying interpersonal relationships. Consequently, in order to compensate for the lack of intimacy and sexuality, in adolescence, the individual would start to rely on autoeroticism and adopt sexual coping for life stressors. During masturbation, the individual would start incorporating deviant sexual fantasies in order to release sexual frustration and compensate for the lack of intimacy. The sexually aggressive tendencies would be further entrenched by the use of cognitive distortions, which could lead to the realisation of sexually violent fantasies if the opportunity arose. Several researchers have tested Marshall's (1989, 1993) theory, and have supported the presence of intimacy deficits and loneliness among sex offenders (Bumby & Hansen, 1997; Martin & Tardif, 2015; Seidman et al., 1994). Additionally, attachment difficulties among sex offenders have been repeatedly found by various researchers (Martin & Tardif, 2015; McKillop et al., 2012; Smallbone & Dadds, 1998; Ward et al., 1996) further supporting the association between attachment styles and sex offending behaviour. 1.2 Sex offenders, and the understanding relationships and intimacy programme Over the years, various treatment approaches have been adopted to reduce sexual recidivism. Most of these treatment programmes follow the principles of cognitive-behavioural therapy (CBT) and relapse prevention, which have been found to have some positive effects on sexual recidivism (Kim et al., 2016). Recently, Grady and colleagues (2016) reported a significant decrease in insecure attachment among incarcerated sex offenders after undergoing a CBT-based treatment programme targeting their specific criminogenic needs. In a review of 48 empirical studies, Jennings and Deming (2017) found that group cohesion, i.e. "the degree to which the group works together, supports and challenges one another" (p. 742), has a highly significant effect on the treatment outcome for the sex offenders, regardless of the treatment modality adopted (e.g. CBT, psychoanalytical). Group cohesion nurtures disclosure and engagement, which in turn reduces denial, which was found to explain 60% of treatment outcome differences (Levenson & Macgowan, 2004; Levenson et al., 2009). Moreover, Lord (2016) also pointed out the benefit of adopting the Good Lives Model (Ward et al., 2007) for enhancing the treatment motivation of sex offenders with mental disorder, as it helps them develop a realistic narrative for personal change leading to greater self-reliance. However, there is limited focus on psychoeducational programmes that aim to improve sex offenders' understanding of relationships and intimacy as means of reducing recidivism. An example of such a programme has been 302 Nusa Crnkovic developed and implemented in a high secure forensic hospital in the UK. The Understanding Relationships and Intimacy (URI) programme is a 20-week group-based psycho-educational programme lasting approximately 2 hours per session on a weekly basis. The URI is intended for individuals who have historically experienced difficulties with interpersonal relationships in one or more social contexts. The different social contexts can be for example within the family, and workplace, including committing an offence. Moreover, the URI programme also targets individuals who have experienced aggression, hostility, loneliness, mistrust or any other form of distress within interpersonal relationships. The overall aim of the URI group programme is to support the development and maintenance of healthy relationships intended for offenders with mental disorder that exhibit problems with establishing and/or maintaining relationships in a prosocial manner. This is done by addressing three key topics: What makes a healthy relationship (ideas of reciprocity, mutual respect and sharing); Identifying healthy sexuality (informed consent, avoiding pressure, combining affection with sexual behaviour); and Increasing knowledge about boundaries and respecting the limits (understanding boundaries in different types of relationship, professional boundaries, consequences of not respecting boundaries). Although this group therapy is primarily psycho-educational, it aims to foster a supportive environment supporting patients' active involvement in the discussions, offering support to other patients or challenging one another on the topics discussed. The current study aims to conduct a preliminary evaluation of the effectiveness of the URI programme specifically for sex offenders, examining its effects on improving their interpersonal relations and intimacy. Based on extant literature it is hypothesised that sex offenders will show less insecure attachment; experience less loneliness; and will report to have improved their interpersonal skills after completing the URI programme. Additionally, their attachment style, levels of self-reported loneliness, and interpersonal skills will be compared to a control group consisting of patients that have not and will not be referred to the URI programme to control for the influence of other factors and programmes offered to the patient in the hospital (e.g. antilibidinal medication, individual therapy). 2 METHODOLOGY 2.1 Sample The study was conducted with male patients at Broadmoor Hospital, a high secure psychiatric treatment facility in the UK for people detained under the Mental Health Act. Participants were chosen based on the patients' responsible clinician's assessment of suitability (e.g. mental health status) to participate in the study and allocated to three groups: (a) sex offender treatment group (SO-TG), (b) non-sex offender treatment group (NSO-TG), and (c) control group (CG). The SO-TG included patients with history of sexual offending behaviour and were participating in the URI programme. The NSO-TG were patients undertaking the URI programme but did not have a history of sexually offending behaviour. The 303 The Effectiveness of a Psycho-Educative Group Programme ... CG comprised of patients who at the time of the study were not participating in the URI programme or completed the same programme in the past. Both of the treatment groups were referred to the URI programme by their multidisciplinary clinical teams based on the patients' past experienced distress in relationships (e.g., aggression, distrust, hostility, loneliness). The sample size used in the study is as follows: SO-TG n = 9, NSO-TG n = 9, and CG n = 10. 2.2 Measures and design All participants completed the following test battery: (a) Inventory of Interpersonal Problems-C [IIP-C] (Horowite et al., 2000), (b) UCLA Loneliness Scale (Russell et al., 1980), and (c) Experience in Close Relationship Scale [ECR-R] (Fraley et al., 2000). The IPP-C (Horowite et al., 2000) is a 32-item self-report questionnaire measuring the presence of interpersonal difficulties. The participants rate the statements on a Likert-type scale ranging from 0 (not at all) to 4 (extremely) based on their belief in how well the short statements describe them. The IIP-C has 8 dimensions measuring potential difficulties one might experience in relations: Dominance (difficulties with aggressive, controlling and/or manipulative behaviour), Intrusiveness (attention seeking, being too open and overly intrusive), Self-Sacrifice (too caring, overly trusting, and attempting too hard to please others), Over Accommodation (exploitive and finds it hard to expressing anger), Non-Assertive (difficulties with being assertive), Social Inhibition (overly socially anxious and inhibited), Coldness (struggles to express emotions and sympathy) and Vindictiveness (being suspicious, distrusting and egocentric) (Barkham et al., 1996). The UCLA Loneliness Scale (Russell et al., 1980) was designed to measure one's subjective experience of loneliness and social isolation. It is a self-report inventory consisting of 20 short statements, which participants rate on a Likert-type scale from 1 (never) to 4 (often). It has high internal consistency (coefficient alpha ranging from 0.86 to 0.94) and test-retest reliability (r = 0.73) over 1-year period (Russell, 1996). The ECR-R (Fraley et al., 2000) is a 36-item self-reported questionnaire measuring adult attachment. The responders are asked to rate short statements with a Likert-like scale ranging from 1 (strongly disagree) to 7 (strongly agree) based on how they generally feel in an intimate relationship. It has two subscales, which measure Anxious and Avoidant attachment styles. The former is characterised by fear of abandonment and rejection, whereas the latter represents avoidance of intimacy and purist of independence. Studies on undergraduates found high internal consistency for both subscales - i.e. for the Anxiety alpha ranges from 0.89 to 0.92, and for the Avoidance subscale, the alpha ranges from 0.91 to 0.95 (Lopez & Gormley, 2002; Wei et al., 2004). Participants were assessed at two time-points, that is, T1 pre- and T2 post-treatment, with a period of 20 weeks between the two test times. Additionally, the nursing staff, who have the most contact with the patients, were asked to fill out the Chart of Interpersonal Relationships in Closed Living Environments 304 Nusa Crnkovic [CIRCLE] (Blackburn & Renwick, 1996) for each patient pre- and post-treatment. The purpose of this was to provide another objective measure of the interpersonal behaviour exhibited by the patients in addition to the self-report questionnaires completed by the patients. However, due to the low numbers of questionnaires being filled out by the nursing staff the data on CIRCLE was subsequently excluded from the study. 2.3 Procedure The list of participants was formed based on their referral to the URI programme and their offence history - i.e. whether or not they committed a sexual offence. Before approaching the patients, permission to do so from their responsible clinicians and team psychologists. All potential concerns regarding any of the potential participants were discussed at the Clinical Team Meeting with the multidisciplinary team [MDT], and patients who were deemed unsuitable to participate (for example, due to mental health deterioration or upcoming move to a different facility before the end of study) were taken off the list. The patients included in the control group were also suggested by the MDTs and permission from their responsible clinicians were obtained before approaching the patients with an invitation to participate in the study. Subsequently, the patients were approached individually and asked to participate in the study which would help determine the effectiveness of the URI programme. The patients were assured that they could decline without any consequences and that their responses would be strictly confidential. It was explained to them that there would be two time-points of testing - the second one being after 20 weeks. They were reassured that everything would be anonymous and if they would wish to withdraw from the study at a later point they can do that again without any consequences. The patients who agreed to participate were first given a consent form, which the researchers explained and checked that the patients fully understood. Subsequently, patients were asked to fill out the ECR-R, the IIP-32 and the UCLA, with researchers being available to answer any potential questions. After the tests were filled out the patients were thanked and told they would be approached once more after 20 weeks for the second testing time. In 20 weeks, the follow-up testing was conducted, with each patient being once more asked whether he still wishes to participate in the study and once more reassured that he can decline to do so without consequences. The primary nurses of each participant were asked to fill out the CIRCLE for the patients also at two time-points. The rationale of the study and the purpose of the CIRCLE was explained to them. After 20 weeks, they were kindly reminded to fill out the CIRCLE again. A positive ethical opinion was obtained for conducting the present study in the NHS (or private sector) by the Nottingham Centre Research Ethics Committee and the West London Mental Health Trust Research and Development Group. 305 The Effectiveness of a Psycho-Educative Group Programme ... 2.4 Data analysis Prior to choosing the statistical tests the normal distribution for the data was first assessed. Although most of the data was normally distributed, several data were nonnormally distributed as well (see Appendix 1). Due to the small sample size, the outliers were not taken out from the data set. Consequently, the non-parametric tests were used for the statistical analysis. Although there were only a few nonnormally distributed results, it has been found that the nonparametric tests' (e.g. Mann-Whitney U test) power is superior to parametric tests' power (e.g. ANCOVA, two-way ANOVA) for small sample sizes when nonnormally distributed results are present (Vickers, 2005). More specifically, the Wilcoxon signed ranks test and the Mann-Whitney U test were used. The Wilcoxon signed ranks test was used to determine whether post-treatment scores of the SO-TG indicate lower levels of loneliness, insecure attachment, and more prosocial interpersonal skills compared to pre-treatment. Subsequently, the Mann-Whitney U test was used to compare the difference in the test results from pre- to post-test of the SO-TG with the difference in the results of the NSO-TG, and the CG. This was used to explore if the emerging changes in the SO-TG were predominantly due to the URI programme rather than other treatment programmes offered at the research site, including anti-libidinal medication. As only two statistical tests were used, no corrections for multiple comparisons were conducted. Nonetheless, this might have resulted in a higher number of erroneous statistically significant results and therefore, interpretation of the statistically significant results must be made with caution. 3 RESULTS 3.1 Participants The average age of the patients involved in the study was 43.45 years (SD = 10.23), ranging from 25 years to 61 years. On average, the participants were in-patients at the high secure psychiatric hospital for 7 years and 10 months (SD = 6.46 years), with the longest stay of 24 years and the shortest stay of 7 months. Predominantly patients were White British (57.1%), followed by 10.7% of Black British, 10.7% of Mixed Race, 3.6% Asian, and 17.6% not identifying with either of the ethnicity categories. There was no significant difference in ethnicity between the SO-TG, NSO-TG and CG (p = 0.26). Based on the ICD-10 (World Health Organisation, 1992), 35.7% (n = 10) of the patients were diagnosed with Dissocial Personality Disorder, 21% (n = 6) with Paranoid Schizophrenia, 10.7% (n = 3) with Specific Personality Disorder, 7.1% (n = 3) with Mixed and Other Personality Disorder, 7.1% (n = 3) with Unspecified Personality Disorder. Additionally, one patient was diagnosed with Emotionally Unstable Personality Disorder, one with Schizophrenia Unspecified, and one with Schizoaffective Disorder. Again, there was no significant difference between the three groups (p = 0.54). 306 Nusa Crnkovic While all of the participants in the SO-TG (n = 9) had a history of sexual offending behaviour and no participants in the NSO-TG (n = 9) had a sexual offence on their criminal record, 60% of the patients (n = 6) in the CG had no history of a sexual offence. No statistically significant difference between the groups was found (p = 0.61). 3.2 Between-subjects change from pre- to post-treatment The differences in scores from the first to second time-point of assessment were calculated for each group. The difference in scores was then statistically analysed by using the Mann-Whitney U test to see if there was a difference between the SO-TG, NSO-TG and CG in the adult attachment style, experiences of loneliness and social isolation, and interpersonal difficulties after completing the URI programme. Using the SPSS version 23, the Mann-Whitney U test was run. The first step of interpreting the result was to examine the histograms to determine whether the data is similarly distributed or not. If the data distribution was not similar, based on which the usage of medians (if the distribution is similar) or mean ranks (if the distribution is not similar) was used for the interpretation of the results. The results are presented in tables 1-3. 3.2.1 Feelings of loneliness and social isolation The loneliness scores for the SO-TG (mean rank = 6.50) and NSO-TG (mean rank = 12.50) were statistically different, U = 13.50, z = -2.39, p = 0.014, R = 0.56. Similarly, the difference in the UCLA scores for the SO-TG (mean rank = 7.11) and CG (mean rank = 12.60) was statistically significant as well, U = 19.00, z = -2.13, p = 0.035, R = 0.49. 3.2.2 Difficulties with interpersonal relationships When exploring the differences between groups, there were no statistically significant results obtained between either of the groups as presented in tables 1-3. Furthermore, there were no significant results found on any of the subscales of the IIP-32 questionnaire when comparing the SO-TG with the NSO-TG or control group. However, one statistically significantly different between the NSO-TG and CG was found in the Vindictiveness dimension was statistically significant, U = 81.00, z = 2.96, p = 0.002, R = 0.68. 3.2.3 Adult attachment style Similarly, when exploring the differences between groups in the attachment style shift from the first to the second time-point of the assessment, no statistically significant results emerged. Neither did the results show a statistically significant difference when comparing the anxious attachment style subscale and the avoidant attachment style subscales between the three groups. 307 The Effectiveness of a Psycho-Educative Group Programme ... Table 1: Between SO-TG and control group comparison of change from pre- to post-treatment SO-TG (n = 9) Control Group (n = 10) Mean Rank Mean Rank MannWhitney U Exact Sig. Z R ECR-R: Total 9.67 10.30 42.00 0.84 -0.25 0.06 ECR-R: Anxious attachment style 9.50 10.45 40.50 0.72 -0.37 0.08 ECR-R: Avoidant attachment style 11.11 9.00 55.00 0.45 0.82 0.19 UCLA 7.11 12.60 19.00 0.04* -2.13 0.49 IIP-32: Social Inhibition 9.94 10.05 44.50 0.97 -0.04 0.01 IIP-32: Non-Assertiveness 8.83 11.05 34.50 0.40 -0.87 0.09 IIP-32: Dominance 9.44 10.50 40.00 0.72 -0.42 0.09 IIP-32: Vindictiveness 10.72 9.35 51.50 0.60 0.53 0.12 IIP-32: Self-Sacrifice 7.67 12.10 24.00 0.09 -1.73 0.39 IIP-32: Coldness 9.61 10.35 41.50 0.78 -0.29 0.07 IIP-32: Intrusive 8.39 11.45 30.50 0.24 -1.19 0.27 IIP-32: Over Accommodation 9.00 10.90 36.00 0.49 -0.76 0.17 IIP-32: Total 9.22 10.70 38.00 0.6 -0.57 0.13 Note: * p < 0.05 Table 2: Between SO-TG and NSO-TG comparison of change from pre- to post-treatment SO-TG (n = 9) NSO-TG (n = 9) Mean Rank Mean Rank MannWhitney U Exact Sig. Z R ECR-R: Total 11.44 7.56 58.00 0.14 1.55 0.36 ECR-R: Anxious attachment style 9.39 9.61 39.50 0.93 -0.09 0.02 ECR-R: Avoidant attachment style 11.22 7.78 56.00 0.19 1.37 0.32 UCLA 6.50 12.50 13.50 0.01** -2.39 0.56 IIP-32: Social Inhibition 8.17 10.83 28.50 0.29 -1.09 0.26 IIP-32: Non-Assertiveness 8.06 10.94 27.50 0.58 -1.174 0.28 IIP-32: Dominance 7.72 11.28 24.50 0.16 -1.42 0.34 IIP-32: Vindictiveness 8.17 10.83 28.50 0.29 1.07 0.08 IIP-32: Self-Sacrifice 9.89 9.11 44.00 0.79 0.32 0.25 IIP-32: Coldness 7.72 11.28 24.50 0.16 -1.43 0.34 IIP-32: Intrusive 7.33 11.67 21.00 0.09 -1.76 0.41 IIP-32: Over Accommodation 9.22 9.78 38.00 0.86 -0.22 0.05 IIP-32: Total 7.17 11.83 19.50 0.06 -1.86 0.44 Note: ** p < 0.01 308 Nusa Crnkovic Table 3: Between NSO-TG and control group comparison of change from pre- to post-treatment ECR-R: Anxious attachment style 9.28 10.65 38.50 0.60 -0.53 0.12 ECR-R: Avoidant attachment style 9.00 10.90 36.00 0.49 -0.74 0.17 UCLA 10.22 9.80 47.00 0.91 0.16 0.68 IIP-32: Social Inhibition 11.50 8.65 58.50 0.28 1.12 0.26 IIP-32: Non-Assertiveness 10.50 9.55 49.50 0.72 0.37 0.09 IIP-32: Dominance 11.28 8.85 56.50 0.36 0.94 0.22 IIP-32: Vindictiveness 14.00 6.40 81.00 0.02* 2.96 0.68 IIP-32: Self-Sacrifice 7.78 12.00 25.00 0.11 -1.64 0.38 IIP-32: Coldness 11.44 8.70 58.00 0.31 1.07 0.25 IIP-32: Intrusive 10.78 9.30 52.00 0.60 0.58 0.25 IIP-32: Over Accommodation 9.61 10.35 41.50 0.78 -0.29 0.07 IIP-32: Total 12.28 7.95 65.50 0.09 1.68 0.38 Note: * p < 0.05 3.3 3.3 Individual improvement from pre- to post-treatment A Wilcoxon signed-rank test was conducted to explore whether there was a significant difference in feelings of loneliness, and difficulties with interpersonal relationships and whether the was a shift from insecure towards more secure attachment after completing the 20-week URI programme in sex offenders and non-sex offenders. 3.3.1 Feelings of loneliness and social isolation The results showed that out of 9 patients with a history of sexual offending, 8 patients reported feeling less lonely and socially isolated (mean rank = 4.5) and one patient reported an increase in the subjective perception of loneliness (mean rank = 9.0). There were no ties. The difference (Mdn = -0.3) in a subjective feeling of loneliness and social isolation was not statistically significant, z = -1.60, p = 0.11, r = 0.38 from pre- (Mdn = 2.6) to post-treatment (Mdn = 2.4). Among NSO-TG the results suggest that out of 9 patients, 5 perceived to be lonelier and socially isolated (mean rank = 5.4) than prior to completion of the URI programme. There was one tie - i.e. there was no change in the experience of loneliness and social isolation, and 3 patients appear to feel less lonely (mean rank = 3.0). The results showed that the difference (Mdn = 0.047) of feelings of loneliness from pre - (Mdn = 1.95) to post-treatment (Mdn = 2.20) was not statistically significant, z = -1.26, p = 0.21, r = 0.29. NSO-TG Control Group (n = 9) (n = 10) Mean Mean Rank Mann- Exact Z R Rank Whitney U Sig. ECR-R: Total 8.11 11.70 28.00 0.18 -1.39 0.32 309 The Effectiveness of a Psycho-Educative Group Programme ... Similarly, out of 10 control participants, 7 reported an increase in feelings of loneliness and social isolation (mean rank = 5.29) and 3 reported feeling lonely less often (mean rank = 6). From pre- (Mdn = 2.25) to post-treatment (Mdn = 2.35) the emerging difference (Mdn = 0.075) was not statistically significant, z = -0.97, p = 0.33, r = 0.22. 3.3.2 Difficulties with interpersonal relationships When exploring the interpersonal difficulties, the results suggested that 5 out of 9 SO-TG patients reported having less overall interpersonal difficulties (mean rank = 5.4) after treatment. There was one tie and 3 patients experienced an increase in interpersonal difficulties (mean rank = 3). The overall difference (Mdn = -0.22) from pre- (Mdn = 1.25) to post-treatment (Mdn = 1.09) was not found to be statistically significant, z = -1.36, p = 0.17, r = 0.32. Neither was there any statistically significant difference from pre- to post-treatment found on the dimensions of Social Inhibition (z = -1.29, p = 0.19, r = 0.30), Non-Assertiveness (z = -1.89, p = 0.058, r = 0.45), Intrusiveness (z = -1.69, p = 0.092, r = 0.39), Self- Sacrifice (z = -1.21, p = 0.23, r = 0.29), and Over Accommodation (z = -1.81, p = 0.071, r = 0.43). However, the results implied that on the Dominance dimension there was a statistically significant decrease in experiencing difficulties with aggression for 6 out of 9 patients (mean rank = 3.5), z = -1.69, p = 0.027, r = 0.52. There were 3 ties as well. Furthermore, 5 out of 9 patients scored lower on the Coldness dimension (mean rank = 3), and for 3 patients there appears to be no change in score. The difference (Mdn = -0.25) from pre- (Mdn = 1.25) to post-treatment (Mdn = 0.75) was found to be statistically significant, z = -2.032, p = 0.042, r = 0.48. Lastly, out of 9 participants, 6 participants were less Vindictive (mean rank = 3.5) and for 3 participants there was no change. The results indicated that the difference (Mdn = -0.5) from pre- (Mdn = 1.25) to post-completion (Mdn = 0) of the URI programme was statistically significant, z = -2.207, p = 0.027, r = 0.52. The analysis of the pre- and post-treatment result for the NSO-TG the results indicated that 7 out of 9 patients experienced more relationship difficulties after attending the URI programme (mean rank = 4.93), whereas for 2 patients there appeared to be a reduction in their interpersonal difficulties (mean rank = 5.25). The difference (Mdn = 0.16) from pre- (Mdn = 0.94) to post-treatment (Mdn = 1.03) was not statistically significant, z = -1.42, p = 0.16, r = 0.34. When examining the dimension, only the Intrusiveness dimension was found to be statistically significant, Z = -2.67, p = 0.008, r = 0.63. More specifically, the results suggest that for all 9 patients there was an increase (Mdn = 1.25) in their difficulties with intrusive behaviour in an interpersonal relationship (mean rank = 5.00) from pre-(Mdn = 0.75) to post-treatment (Mdn = 2.00). For the remaining 7 dimensions the change from pre- to post-treatment was found to be statistically nonsignificant -i.e. Social Inhibition (z = -0.43, p = 0.67, r = 0.10), Non-Assertiveness (z = -0.64, p = 0.52, r = 0.15), Dominance (z = -0.12, p = 0.91, r = 0.028), Self-Sacrifice (z = -0.95, p = 0.34, r = 0.22), Coldness (z = -0.77, p = 0.44, r = 0.18), Vindictiveness (z = -0.21, p = 0.83, r = 0.049) and Over Accommodation (z = -1.40, p = 0.16, r = 0.33). Similar results were also obtained for the control group where overall 7 out of 10 patients had fewer interpersonal difficulties at the second testing time (mean 310 Nusa Crnkovic rank = 5.50). On the other hand, results suggested that 3 patients experienced more interpersonal issues (mean rank = 5.5). Nonetheless, the difference (Mdn = -0.17) between first (Mdn = 1.11) and second (Mdn = 1.05) time-point was statistically not significant, z = -1.12, p = 0.26, r = 0.25. There was no statistically significant change from first assessment to second assessment point on 7 of the interpersonal difficulties dimensions - i.e. Social Inhibition (z = -0.72, p = 0.46, r = 0.16), Non-Assertiveness (z = -1.34, p = 0.18, r = 0.29), Dominance (z = -1.19, p = 0.23, r = 0.27), Coldness (z = -0.92, p = 0.36, r = 0.20), Vindictiveness (z = -1.49, p = 0.14, r = 0.33) and Over Accommodation (z = -1.38, p = 0.17, r = 0.31). For the Self-Sacrifice dimension there was a statistically different (Mdn = 1.25) result found between the first point of time (Mdn = 0.88) and second point of time (Mdn = 2.50), z = -2.65, p = 0.008, r = 0.33. 3.3.3 Adult attachment style The Wilcoxon signed-rank test showed that for 3 out of 9 SO-TG participants there was no change in the attachment style. For 4 of the patients their results implied that there was an increase in an insecure attachment style (mean rank = 3.5), whereas for 2 patients there appears to be a change towards a more secure attachment style after the URI programme (mean rank = 3.5). However, the difference (Mdn = 0) from pre- (Mdn = 4.53) to post-treatment (Mdn = 4.67) was statistically insignificant, z = -0.73, p = 0.46, r = 0.17. When exploring the change in Anxious attachment style, the results suggest that for 4 out of 9 patients there was a decrease in their insecure attachment (mean rank = 5.50), for one patient there was no change, and for the remaining 4 there was an increase in their Anxious attachment style (mean rank = 3.50). Neither of the changes were statistically significant - i.e. Anxious attachment style (z = -0.56, p = 0.58, r = 0.13) and Avoidant attachment style (z = -0.98, p = 0.33, r = 0.23) attachment style. For the NSO-TG results showed that for 7 out of 9 patients there was a shift from insecure attachment towards a more secure attachment style (mean rank = 4.79). The results for two patients indicated an increase in an insecure attachment style (mean rank = 5.75). The difference (Mdn = -0.22) in insecure attachment from pre- (Mdn = 4.47) to post-treatment (Mdn = 4.36) was statistically not significant, z = -1.30, p = 0.19, r = 0.24. Out of 9 participants, 6 patients had a lower Anxious attachment style after treatment (mean rank = 4.50), although 3 participants exhibited an increase in Anxious attachment style (mean rank = 6.00). Likewise, there was no statistically significant change on the subscales measuring the Anxious attachment style (z = -0.53, p = 0.59, r = 0.31) and Avoidant attachment style (z = -1.01, p = 0.31, r = 0.13). When comparing the attachment style change in CG after 20 weeks, the results suggest that insecure attachment increased for 6 out of 10 (mean rank = 5.83) and decreased for the other 4 patients (mean rank = 5.00). The difference (Mdn = 0.75) between levels of insecure attachment at the first time-point of testing (Mdn = 3.96) and at the second time-point of testing (Mdn = 4.22) was statistically nonsignificant, z = -0.77, p = 0.44, r = 0.057. Furthermore, for half of the patients, there was an increase in Anxious attachment style (mean rank = 5.30) and for the other half, there was a decrease in Anxious attachment style (mean rank = 5.70). Similarly, 311 The Effectiveness of a Psycho-Educative Group Programme ... there was a decrease in Avoidant attachment style for 6 patients (mean rank = 4.17) and an increase for the remaining 4 patients (mean rank = 7.50). The change was not statistically significant for the Anxious attachment style (z = -0.10, p = 0.92, r = 0.17) nor for the Avoidant attachment style (z = -0.26, p = 0.79, r = 0.023). 4 DISCUSSION The present study aimed to conduct a preliminary evaluation of the effectiveness of the Understanding Relationships and Intimacy (URI) psycho-educative group therapy for sex offenders. It was hypothesised that after completion of the URI programme the sex offenders would report significantly fewer experiences of interpersonal difficulties, feelings of loneliness and social isolation, and would become more securely attached. Thus, if the results would support the hypothesis, it would imply that the URI programme has a beneficial effect on sex offenders' understanding of intimacy and relationships, which might in turn reduce their future recidivism risk. The results showed that after completing the URI psycho-educative group therapy, SO-TG experienced significantly fewer feelings of loneliness and social isolation compared to the control group. Furthermore, SO-TG reported fewer feelings of loneliness and social isolation in comparison to NSO-TG who also undertook the URI treatment programme. The reduction in loneliness could be because the URI programme is being conducted in a group setting, which was found to be a preferred therapeutic setting by sex offenders (Levenson et al., 2014). Sex offenders, especially child molesters, tend to be highly stigmatised, not only by the general public but also within the criminal system, which promotes their social isolation and loneliness (Ferguson & Ireland, 2006; Ricciardelli & Moir, 2013; Tewksbury, 2012). Group therapy setting, on the other hand, might be the only social setting in which sex offenders disclose their stigmatisation and feel supported (Frost & Connolly, 2004; Jennings & Deming, 2017), thus, reducing their feeling of loneliness and social isolation. Moreover, NSO-TG also reported feeling less lonely after completing the URI programme compared to the control group, which supports the importance of group cohesion on treatment outcomes as pointed out by Jennings and Deming (2017). However, when examining the within-subject changes in the subjective experience of loneliness and social isolation, the results for none of the three groups of participants were statistically significant. One possible explanation could be that no change was found due to the small sample size and the use of only self-reported questionnaires. The effect sizes for all three groups were small, allowing the possibility that there might have been different results found if the sample size would be bigger. An alternative explanation could be the different criminogenic needs of the URI group therapy participants. More specifically, the patients involved in the URI programme had different ICD-10 diagnoses, varying from personality disorders to schizophrenic disorders, and offending history. The high diversity of the patients involved in the group therapy might have impaired the Need and Responsivity factors of the Risk, Need, Responsivity (RNR) Model (Andrews & Bonta, 1998) by requiring excessive flexibility in the programme, yet 312 Nusa Crnkovic failing to adapt sufficiently to different needs patients with different diagnosis present with and consequently, reducing the treatment effectiveness. Additionally, because the URI programme is psycho-educational, it might not foster group cohesion and expressiveness of one's affects sufficiently enough. When comparing the presence of interpersonal difficulties between the SG-TG and NSO-TG who completed the URI group therapy, there was no statistically significant difference found. Similarly, no differences were found when comparing the SG-TG to the control group and the NSO-TG to the control group. All of the effect sizes were small to medium, which implies that if the study were to be conducted on a larger sample size, there would be a statistically significant difference. However, when examining the effect sizes, there was a medium effect size when comparing the SO-TG with NSO-TG and NSO-TG with CG. On the other hand, when comparing SO-TG with the control group the effect size was small. This suggests that SO-TG had more interpersonal difficulties before completing the URI group than did the NSO-TG, and thus had a greater reduction in their interpersonal difficulties after undertaking the programme. This notion is also supported by analysing the within-subject difference from pre- to post-treatment. More specifically, although once more the results were statistically not significant, when examining the mean ranks of SO-TG and NSO-TG it can be seen that there was a greater reduction for SO-TG than for the NSO-TG. Moreover, the median difference for the SO-TG implied a reduction in interpersonal difficulties from pre- to post-treatment, whereas the median difference from pre- to post-treatment for NSO-TG implied an increase in interpersonal difficulties. The decrease in interpersonal difficulties for the SO-TG and an increase in interpersonal difficulties for NSO-TG might again reflect the aforementioned high diversity of the patients in the URI programme, again supporting the proposition that the URI programme did not efficiently adopt the need and responsivity factors of the RNR (Andrews & Bonta, 1998) model. Further exploration of SO-TG interpersonal difficulties was made by examining the eight dimensions of interpersonal issues, namely: dominance, intrusiveness, self-sacrifice, over accommodation, non-assertiveness, social inhibition, coldness and vindictiveness. The results showed that there was no statistically significant difference between SO-TG and NSO-TG or the control group on any of the dimensions. However, there was a statistically significant difference with a strong effect size between NSO-TG and CG on the Vindictiveness dimension. Yet when examining the within-subject difference from the first to the second point of assessment on the Vindictiveness dimension, there was no statistical difference for NSO-TG or CG although the effect sizes were small. Interestingly, there was a statistically significant difference with medium effect size on the Vindictiveness dimension for SO-TG. This suggests that SO-TG compared to NSO-TG and CG had more interpersonal difficulties associated with being egocentric, suspicious and distrusting. Moreover, a statistically significant decrease from pre- to postcompetition of the URI programme in the Dominance and Coldness dimensions was also found in SO-TG. Although there was no statistically significant difference found for the Dominance or Coldness dimensions in the NSO-TG, there was a statistically significant reduction in the Intrusiveness dimension. 313 The Effectiveness of a Psycho-Educative Group Programme ... The diverse findings on the URI group's impact on interpersonal difficulties further support the suggestion that the URI programme still has room for improvement with the implementation of the RNR model (Andrews & Bonta, 1998). Moreover, as the URI is a psycho-educational programme with a more structured agenda this might hinder open expression of affects for all of the patients and group cohesion, the latter being found to be highly important for the treatment outcome for the sex offenders (Jennings & Deming, 2017). Lastly, the effect of the URI group on the adult attachment styles was explored. There was no statistical difference in overall adult attachment style when comparing the SO-TG with the NSO-TG or the control group. Neither was a statistically significant difference found on the subscales of Anxious attachment style or Avoidant attachment style. Similarly, there was no statistically significant change from pre- to post-treatment in attachment styles in the SO-TG or the NSO-TG. One possible explanation for this is the difference in attachment styles between different types of sex offenders. While rapists tend to exhibit a more avoidant attachment style, child molesters tend to have an anxious attachment style, originating from different childhood experiences and consequently leading to different interpersonal difficulties in adulthood (Martin & Tardif, 2014; McCormack et al., 2002; Ward et al., 1996). Therefore, the lack of statistically significant results cannot be solely contributed to the URI programme, but may also reflect the small sample size used in the study which prevented further exploration of the attachment styles among different types of sex offenders. Taken all together, the results of the present research suggest that the URI programme fails to effectively achieve its own primary aim for participants with sex-offending history - i.e. to support the development and maintenance of healthy relationships among offenders with mental disorder that exhibit problems with establishing and/or maintaining relationships in a prosocial manner. The manner in which one engages in interpersonal relationships is significantly more complex than solely possessing a rational understanding on »how they work«. It is affected by the person's attachment style, which is a complex reflection of early experiences with primary caregiver, social competences and quality of their close relationships (Bowlby, 1973; Fraley et al., 2013; Ozcan et al., 2016). Numerous researchers have found significantly high levels of childhood trauma among sex offenders (e.g. Levenson et al., 2014; McCormack et al., 2002; Seghorn et al., 1987), which is also closely linked to insecure attachment styles (e.g. Maniglio, 2011; Ozcan et al., 2016). Therefore, with focusing predominately solely on the cognitive aspect of the social skills the URI programme appears to neglect the need and responsivity aspect of the RNR model by underestimating the complex interpersonal dynamics needed for formation and retention of healthy interpersonal relationships. 4.1 Limitations The current study has several limitations, the most crucial being the small sample size. The small number of participants in the study is a consequence of the patients either not wishing to participate or dropping out at the second time-point of assessment for various reasons (e.g. discharged to a different facility, did not »feel 314 Nusa Crnkovic like it«, mental health deterioration). This prevented a more detailed examination of the URI's effectiveness on different types of sex offenders. Furthermore, the small sample is also partly reflective of the fact that there was a 20-week waiting period in between two time-points of testing and that there was a time constraint on the research, limiting the possibility of expanding the sample size further. Moreover, as pointed out in the discussion section, because of a small sample the question remains whether the statistically insignificant results represent the actual absence of improvement on the tested constructs or whether a different trend would emerge if a bigger sample size would be at the disposal. The additional limitation is the absence of objective measurements of patients' interpersonal relations. Although the initial study design attempted to include the CIRCLE (Blackburn & Renwick, 1996) it had to be taken out because it was completed only for a few participants. Given that the sample size was very small, the only option was to exclude CIRCLE rather than to exclude the patients for whom the CIRCLE was not filled out. Thus, the entire study is based on self-reported questionnaires without evaluating the validity of patients' responses, leaving the possibility of the results not being representative of the patients' genuine attachment styles, feelings of loneliness and difficulties with intimacy. Additionally, results would be more informative if a wait-list for the URI programme group would be included. That is if we could compare patients who have completed the URI programme already and patients who have been identified that would benefit from the URI programme but have not yet completed it. Lastly, in order to see whether the URI programme is indeed effective, another follow-up testing after 20 weeks should be conducted in order to assess whether the emerging changes are stable over time. This would make the assessment of the URI group's effectiveness significantly more representative and valid. 5 CONCLUSIONS AND FUTURE RESEARCH The URI programme was effective in reducing the subjective experience of loneliness when comparing sex offenders that have completed the URI group with the non-sex offenders that completed the URI group and the control group. Furthermore, it reduced SO-TG's egocentrism, suspiciousness and distrustfulness. Overall, URI treatment is not effectively targeting sex offenders' insecure attachment styles, interpersonal difficulties, or feelings of loneliness and social isolation. This appears to be due to the overly diverse treatment group in terms of the ICD-10 diagnoses and the offending history of the patients, reducing the ability to successfully apply the RNR model to the group therapy. Although the URI is not specifically intended for sex offenders, one possible solution might be to form groups based on their attachment styles. That would enable them to adapt the URI programme more according to the RNR model. Future research should strive to repeat the current study on bigger sample size (e. g. N = 100 per group), including the treatment group, wait-list group and control group in order to assess the effectiveness of the URI programme. The sex offenders involved in the study should be split into two different groups - i.e. rapists and child molesters, in order to see whether there are indeed different 315 The Effectiveness of a Psycho-Educative Group Programme ... emerging trends between these two types of sexual offenders and if so, what are the differences and what are the similarities. It should include observational psychometric instruments (e.g. CIRCLE), and interviews, which would allow a more in-depth exploration of the patients' attachment style, difficulties with interpersonal relationships and intimacy, and feelings of loneliness. Additionally, it should include a third time-point of testing to determine the perseverance of change which might result from the URI programme in order to obtain a genuine reflection of the effectiveness of the URI programme for sex offenders. Acknowledgments: I would like to express my gratitude to my mentors dr. James Tapp, dr. Alex Lord and prof. dr. Derek Perkins for enabling me to conduct the study at the hospital, and for all the guidance and support during the process. REFERENCES Ainsworth, M. D. S., & Bell, S. M. (1981). Attachment, exploration, and separation: Illustrated by the behavior of one-year-olds in a strange situation. In The life cycle (pp. 57-71). Columbia University Press. Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. N. (2015). Patterns of attachment: A psychological study of the strange situation. Psychology Press. Andrews, D. A., & Bonta, J. (1998). The psychology of criminal conduct (2nd ed.). Anderson Publishing. Armstrong, J., & Mellor, D. (2016). Internet child pornography offenders: An examination of attachment and intimacy deficits. Legal and Criminological Psychology, 21(1), 41-55. https://doi.org/10.1111/lcrp.12028 Blackburn, R., & Renwick, S. J. (1996). Rating scales for measuring the interpersonal circle in forensic psychiatric patients. Psychological Assessment, 8(1), 76-84. https://doi.org/10.1037/1040-3590.8.1.76 Bowlby, J. (1973). Attachment and loss [Vol. II: Separation: Anxiety and anger]. Basic Books. Barkham, M., Hardy, G. E., & Startup, M. (1996). The IIP-32: A short version of the inventory of interpersonal problems. British Journal of Clinical Psychology, 35(1), 21-35. https://doi.org/10.1111/j.2044-8260.1996.tb01159.x Bumby, K. M., & Hansen, D. J. (1997). Intimacy deficits, fear of intimacy, and loneliness among sexual offenders. Criminal Justice and Behavior, 24(3), 315331. https://doi.org/10.1177/0093854897024003001 Carlson, E. A., & Ruiz, S. K. (2016). Transactional processes in the development of adult personality disorder symptoms. Development and Psychopathology, 28(3), 639-651. https://doi.org/10.1017/S0954579416000225 Carr, S. C., Hardy, A., & Fornells-Ambrojo, M. (2018). Relationship between attachment style and symptom severity across the psychosis spectrum: A meta-analysis. Clinical Psychology Review, 59, 145-158. https://doi.org/10.1016/j. cpr.2017.12.001 Cortoni, F., & Marshall, W. L. (2001). Sex as a coping strategy and its relationship to juvenile sexual history and intimacy in sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 13(1), 27-43. https://doi. org/10.1177/107906320101300104 316 Nusa Crnkovic Craissati, J. (2009). Attachment problems and sex offending. In A. R. Beech, L. A. Craig, & K. D. Browne, (Eds.), Assessment and treatment of sex offenders: A handbook (pp. 18-38). J. Wiley. Cuoco, S., Nistico, V., Cappiello, A., Scannapieco, S., Gambini, O., Barone, P., Erro, R., & Demartini, B. (2021). Attachment styles, identification of feelings and psychiatric symptoms in functional neurological disorders. Journal of Psychosomatic Research, 147, 110539. https://doi.org/10.1016/j. jpsychores.2021.110539 Ferguson, K., & Ireland, C. (2006). Attitudes towards sex offenders and the influence of offence type: A comparison of staff working in a forensic setting and students. The British Journal of Forensic Practice, 8(2), 10-19. https://doi. org/10.1108/14636646200600009 Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78(2), 350-365. https://doi.org/10.1037/0022-3514.78.2.350 Fraley, R. C., Roisman, G. I., Booth-LaForce, C., Owen, M. T., & Holland, A. S. (2013). Interpersonal and genetic origins of adult attachment styles: A longitudinal study from infancy to early adulthood. Journal of Personality and Social Psychology, 104(5), 817-838. https://doi.org/10.1037/a0031435 Frost, A., & Connolly, M. (2004). Reflexivity, reflection, and the change process in offender work. Sexual Abuse, 16(4), 365-380. https://doi.org/10.1023/ B:SEBU.0000043329.65730.ff Grady, M. D., Swett, L., & Shields, J. J. (2016). The impact of a sex offender treatment programme on the attachment styles of incarcerated male sexual offenders. Journal of Sexual Aggression, 22(1), 123-136. https://doi.org/10.1080/ 13552600.2014.894148 Grady, M. D., Yoder, J., & Brown, A. (2018). Childhood maltreatment experiences, attachment, sexual offending: Testing a theory. Journal of Interpersonal Violence, 36(11-12), NP6183-NP6217. https://doi.org/10.1177/0886260518814262 Groh, A. M., Fearon, R. P., Van IJzendoorn, M. H., Bakermans-Kranenburg, M. J., & Roisman, G. I. (2017). Attachment in the early life course: Meta-analytic evidence for its role in socioemotional development. Child Development Perspectives, 11(1), 70-76. https://doi.org/10.1111/cdep.12213 Horowite, L. M., Alden, L. E., Wiggins, J. S., & Pincus, A. L. (2000). Inventory of interpersonal problems (IIP-32/IIP-64). Psychological Corporation. Jennings, J. L., & Deming, A. (2017). Review of the empirical and clinical support for group therapy specific to sexual abusers. Sexual Abuse, 29(8), 731-764. https://doi.org/10.1177/1079063215618376 Jespersen, A. F., Lalumiere, M. L., & Seto, M. C. (2009). Sexual abuse history among adult sex offenders and non-sex offenders: A meta-analysis. Child abuse & neglect, 33(3), 179-192. https://doi.org/10.1016/j.chiabu.2008.07.004 Kim, B., Benekos, P. J., & Merlo, A. V. (2016). Sex offender recidivism revisited: Review of recent meta-analyses on the effects of sex offender treatment. Trauma, Violence, & Abuse, 17(1), 105-117. https://doi.org/10.1177/1524838014566719 Levenson, J. S., & Macgowan, M. J. (2004). Engagement, denial, and treatment progressamongsexoffenders ingrouptherapy. SexualAbuse:AJournalofResearch 317 The Effectiveness of a Psycho-Educative Group Programme ... and Treatment, 16(1), 49-63. https://doi.org/10.1177/107906320401600104 Levenson, J. S., Macgowan, M. J., Morin, J. W., & Cotter, L. P. (2009). Perceptions of sex offenders about treatment: Satisfaction and engagement in group therapy. Sexual Abuse, 21(1), 35-56. https://psycnet.apa.org/ doi/10.1177/1079063208326072 Levenson, J. S., Prescott, D. S., & Jumper, S. (2014). A consumer satisfaction survey of civilly committed sex offenders in Illinois. International journal of offender therapy and comparative criminology, 58(4), 474-495. https://doi. org/10.1177/0306624X12472956 Levenson, J. S., Willis, G. M., & Prescott, D. S. (2015). Adverse childhood experiences in the lives of female sex offenders. Sexual Abuse, 27(3), 258-283. https://doi.org/10.1177/1079063214544332 Levinson, A., & Fonagy, P. (2004). Offending and attachment: The relationship between interpersonal awareness and offending in a prison population with psychiatric disorder. Canadian Journal of Psychoanalysis, 12(2), 225-251. Lopez, F. G., & Gormley, B. (2002). Stability and change in adult attachment style over the first-year college transition: Relations to self-confidence, coping, and distress patterns. Journal of Counseling Psychology, 49(3), 355-364. https://doi. org/10.1037/0022-0167.49.3.355 Lord, A. (2016). Integrating risk, the Good Lives Model and recovery for mentally disordered sexual offenders. Journal of Sexual Aggression, 22(1), 107-122. https://doi.org/10.1080/13552600.2014.975164 Main, M., & Solomon, J. (1990). Procedures for identifying infants as disorganized/ disoriented during the Ainsworth Strange Situation. In M. T. Greenberg, D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years: Theory, research, and intervention (pp. 121-160). The University of Chicago Press. Maniglio, R. (2011). The role of childhood trauma, psychological problems, and coping in the development of deviant sexual fantasies in sexual offenders. Clinical Psychology Review, 31(5), 748-756. https://doi.org/10.1016/j. cpr.2011.03.003 Marshall, W. L. (1989). Intimacy, loneliness and sexual offenders. Behaviour Research and Therapy, 27(5), 491-503. https://doi.org/10.1016/0005-7967(89)90083-1 Marshall, W. L. (1993). The role of attachments, intimacy, and loneliness in the etiology and maintenance of sexual offending. Sexual and Relationship Therapy, 25(1), 73-85. http://dx.doi.org/10.1080/14681990903550191 Marshall, W. L., & Marshall, L. E. (2000). The origins of sexual offending. Trauma, Violence, & Abuse, 1(3), 250-263. https://doi.org/10.1177/1524838000001003003 Marshall, W. L., Marshall, L. E., Serran, G. A., & O'Brien, M. D. (2009). Self-esteem, shame, cognitive distortions and empathy in sexual offenders: Their integration and treatment implications. Psychology, Crime & Law, 15(2-3), 217234. https://doi.org/10.1080/10683160802190947 Martin, G. M., & Tardif, M. (2014). What we do and don't know about sex offenders' intimacy dispositions. Aggression and Violent Behavior, 19(4), 372-382. https:// doi.org/10.1016/j.avb.2014.06.002 Martin, G. M., & Tardif, M. (2015). Examining sex offenders' intimacy deficits: Their nature and their influence on sexually abusive behaviours. Journal of 318 Nusa Crnkovic Sexual Aggression, 21(2), 158-178. https://doi.org/1Q.1Q80/135526QQ.2Q13.84976 8 McCormack, J., Hudson, S. M., & Ward, T. (2QQ2). Sexual offenders' perceptions of their early interpersonal relationships: An attachment perspective. Journal of Sex Research, 39(2), 85-93. https://doi.org/1Q.1Q8Q/QQ22449Q2Q9552127 McKillop, N., Smallbone, S., Wortley, R., & Andjic, I. (2Q12). Offenders' attachment and sexual abuse onset: A test of theoretical propositions. Sexual Abuse, 24(6), 591-61Q. https://doi.org/1Q.1177/1Q79Q63212445571 Meloy, J. R., & Gothard, S. (1995). Demographic and clinical comparison of obsessional followers and offenders with mental disorders. The American Journal of Psychiatry, 152(2), 258-263. https://doi.org/1Q.1176/ajp.152.2.258 Mikulincer, M., & Shaver, P. R. (2QQ7). Attachment in adulthood: Structure, dynamics, and change. The Guilford Press. Ogilvie, C. A., Newman, E., Todd, L., & Peck, D. (2Q14). Attachment & violent offending: A meta-analysis. Aggression and Violent Behavior, 19(4), 322-339. https://doi.org/1Q.1Q16/j.avb.2Q14.Q4.QQ7 Ozcan, N. K., Boyacioglu, N. E., Enginkaya, S., Bilgin, H., & Tomruk, N. B. (2Q16). The relationship between attachment styles and childhood trauma: A transgenerational perspective - A controlled study of patients with psychiatric disorders. Journal of Clinical Nursing, 25(15-16), 2357-2366. https:// doi.org/1Q.1111/jocn.13274 Popovic, M. (2QQ5). Intimacy and its relevance in human functioning. Sexual and Relationship Therapy, 20(1), 31-49. https://doi.org/1Q.1Q8Q/1468199Q412331323 992 Rice, M. E., & Harris, G. T. (1997). The treatment of mentally disordered offenders. Psychology, Public Policy, and Law, 3(1), 126 -183. https://doi.org/1Q.1Q37/1Q76-8971.3.1.126 Ricciardelli, R., & Moir, M. (2Q13). Stigmatized among the stigmatized: Sex offenders in Canadian penitentiaries. Canadian Journal of Criminology and Criminal Justice, 55(3), 353-386. https://doi.org/1Q.3138/cjccj.2Q12.E22 Russell, D., Peplau, L. A., & Cutrona, C. E. (198Q). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472-48Q. https://doi.org/1Q.1Q37/QQ22-3514.39.3.472. Russell, D. W. (1996). UCLA loneliness scale (Version 3): Reliability, validity, and factor structure. Journal of Personality Assessment, 66(1), 2Q-4Q. https://doi. org/1Q.12Q7/s15327752jpa66Q1 2 Seghorn, T. K., Prentky, R. A., & Boucher, R. J. (1987). Childhood sexual abuse in the lives of sexually aggressive offenders. Journal of the American Academy of Child & Adolescent Psychiatry, 26(2), 262-267. https://doi.org/1Q.1Q97/QQQQ4583-1987Q3QQQ-QQQ25 Seidman, B. T., Marshall, W. L., Hudson, S. M., & Robertson, P. J. (1994). An examination of intimacy and loneliness in sex offenders. Journal of Interpersonal Violence, 9(4), 518-534. https://doi.org/1Q.1177/Q88626Q94QQ9QQ4QQ6 Smallbone, S. W., & Dadds, M. R. (1998). Childhood attachment and adult attachment in incarcerated adult male sex offenders. Journal of Interpersonal Violence, 13(5), 555-573. https://doi.org/1Q.1177/Q88626Q98Q13QQ5QQ 319 The Effectiveness of a Psycho-Educative Group Programme ... Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, W. A. (2009). The development of the person: The Minnesota study of risk and adaptation from birth to adulthood. Guilford Press. Tewksbury, R. (2012). Stigmatization of sex offenders. Deviant Behavior, 33(8), 606623. https://doi.org/10.1080/01639625.2011.636690 Thornton, D. (2002). Constructing and testing a framework for dynamic risk assessment. Sexual Abuse, 14(2), 139-153. https://doi. org/10.1177/107906320201400205 Van IJzendoorn, M. H., Feldbrugge, J. T. T. M., Derks, F. C. H., De Ruiter, C., Verhagen, M. F. M., Philipse, M. W. G., Van der Staak, C. P. F., & Riksen-Walraven, J. M. A. (1997). Attachment representations of personality-disordered criminal offenders. American Journal of Orthopsychiatry, 67(3), 449459. https://doi.org/10.1037/h0080246 Vickers, A. J. (2005). Parametric versus non-parametric statistics in the analysis of randomized trials with non-normally distributed data. BMC Medical Research Methodology, 5(1), 35. https://doi.org/10.1186/1471-2288-5-35 Waldinger, R. J., Schulz, M. S., Barsky, A. J., & Ahern, D. K. (2006). Mapping the road from childhood trauma to adult somatization: The role of attachment. Psychosomatic Medicine, 68(1), 129-135. https://doi.org/10.1097/01. psy.0000195834.37094.a4 Ward, T., Hudson, S. M., & Marshall, W. L. (1996). Attachment style in sex offenders: A preliminary study. Journal of Sex Research, 33(1), 17-26. https:// doi.org/10.1080/00224499609551811 Ward, T., McCormack, J., & Hudson, S. M. (1997). Sexual offenders' perceptions of their intimate relationships. Sexual Abuse: Journal of Research and Treatment, 9(1), 57-74. https://doi.org/10.1177/107906329700900105 Ward, T., Mann, R. E., & Gannon, T. A. (2007). The good lives model of offender rehabilitation: Clinical implications. Aggression and Violent Behavior, 12(1), 87107. https://doi.org/10.1016/j.avb.2006.03.004 Wei, M., Mallinckrodt, B., Russell, D. W., & Abraham, W. T. (2004). Maladaptive perfectionism as a mediator and moderator between adult attachment and depressive mood. Journal of Counselling Psychology, 51(2), 201-212. https://doi. org/10.1037/0022-0167.51.2.201 Wynne, L. C., & Wynne, A. R. (1986). The quest for intimacy. Journal of Marital and Family Therapy, 12(4), 383-394. https://doi.org/10.1111/j.1752-0606.1986. tb00671.x About the Author: Nusa Crnkovic, MSc in forensic psychology, development & research associate in the field of mental health at National Institute of Public Health. E-mail: Nusa.crnkovic@nijz.si 320 Nusa Crnkovic APPENDIX Appendix 1 Table 4: Normal distribution assessment of the Sex offending treatment group data N Mean Std. Deviation Skewness Kurtosis Zs Zk Variables Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error Pre- ECR-R: Total 9 4.47 0.64 -0.10 0.72 -0.98 1.40 -0.15 -0.69 Post- ECR-R: Total 9 4.63 0.39 0.19 0.72 0.51 1.40 0.28 0.36 Pre- ECR-R: Anxious attachment style 9 4.27 0.39 1.06 0.72 0.69 1.40 1.49 0.49 Post- ECR-R: Anxious attachment style 9 4.17 0.41 -1.12 0.72 2.10 1.40 -1.56 1.50 Pre- ECR-R: Avoidant attachment style 9 4.69 1.12 0.16 0.72 -1.48 1.40 0.23 -1.05 Post- ECR-R: Avoidant attachment style 9 5.09 0.69 0.18 0.72 0.24 1.40 0.26 0.17 Pre- UCLA 9 2.48 0.58 -1.60 0.72 3.46 1.40 -2.24 2.47 Post- UCLA 9 2.28 0.31 -0.90 0.72 -0.17 1.40 -1.27 -0.12 Pre- IIP-32: Social Inhibition 9 1.14 0.85 -0.15 0.72 -1.39 1.40 -0.21 -0.99 Post- IIP-32: Social Inhibition 9 0.94 0.77 0.87 0.72 0.98 1.40 1.21 0.70 Pre- IIP-32: Non-Assertiveness 9 1.72 1.61 0.38 0.72 -1.71 1.40 0.53 -1.22 Post- IIP-32: Non-Assertiveness 9 1.0 0.97 0.54 0.72 -1.44 1.40 0.76 -1.03 Pre- IIP-32: Dominance 9 1.53 1.48 0.45 0.72 -1.19 1.40 0.64 -0.85 Post- IIP-32: Dominance 9 0.89 1.09 1.11 0.72 0.19 1.40 1.55 0.14 Pre- IIP-32: Vindictiveness 9 1.33 0.98 0.67 0.72 0.69 1.40 0.93 0.49 Post- IIP-32: Vindictiveness 9 2.03 0.64 0.01 0.72 -0.67 1.40 0.02 -0.48 Pre- IIP-32: Self-Sacrifice 9 0.53 0.59 0.48 0.72 -1.61 1.40 0.68 -1.15 Post- IIP-32: Self-Sacrifice 9 1.11 1.54 1.09 0.72 -0.22 1.40 1.53 -0.16 321 The Effectiveness of a Psycho-Educative Group Programme ... Pre- IIP-32: Coldness 9 1.33 1.15 0.42 0.72 -1.15 1.40 0.59 -0.82 Post- IIP-32: Coldness 9 0.97 0.99 1.16 0.72 0.92 1.40 1.62 0.66 Pre- IIP-32: Intrusive 9 1.36 1.47 0.74 0.72 -0.65 1.40 1.03 -0.47 Post- IIP-32: Intrusive 9 0.53 0.71 1.19 0.72 0.84 1.40 1.66 0.59 Pre- IIP-32: Over Accommodation 9 1.64 1.39 0.73 0.72 -0.71 1.40 1.02 -0.51 Post- IIP-32: Over Accommodation 9 1.17 1.02 0.08 0.72 -1.55 1.40 0.12 -1.11 Pre- IIP-32: Total 9 1.32 1.09 0.44 0.72 -0.82 1.40 0.62 -0.59 Post- IIP-32: Total 9 0.98 0.65 0.98 0.72 1.02 1.40 1.37 0.73 Valid N (listwise) 9 Note: The cut-off score Z = 1.64 was used to determine whether or not the data is normally distributed. Nonnormally distributed data is marked in bold. Table 5: Normal distribution assessment of the non -sex offending treatment group data N Mean Std. Deviation Skewness Kurtosis Zs Zk Variables Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error Pre- ECR-R: Total 9 4.29 0.50 -0.47 0.72 -1.24 1.40 -0.66 -0.88 Post- ECR-R: Total 9 4.14 0.59 -1.19 0.72 1.19 1.40 -1.67 0.85 Pre- ECR-R: Anxious attachment style 9 3.81 0.47 -0.63 0.72 -0.21 1.40 -0.87 -0.15 Post- ECR-R: Anxious attachment style 9 3.74 0.52 -0.40 0.72 0.24 1.40 -0.56 0.17 Pre- ECR-R: Avoidant attachment style 9 4.53 0.98 -0.25 0.72 -1.29 1.40 -0.35 -0.92 Post- ECR-R: Avoidant attachment style 9 4.77 0.88 0.19 0.72 -1.50 1.40 0.26 -1.07 Pre- UCLA 9 2.05 0.52 0.59 0.72 -0.96 1.40 0.83 -0.69 Post- UCLA 9 2.21 0.47 -0.60 0.72 1.50 1.40 -0.84 1.07 Pre- IIP-32: Social Inhibition 9 1.19 0.74 0.81 0.72 2.64 1.40 1.12 1.89 Post- IIP-32: Social Inhibition 9 1.39 0.76 0.55 0.72 -1.32 1.40 0.76 -0.94 322 Nusa Crnkovic Pre- IIP-32: Non-Assertiveness 9 1.19 0.72 -0.25 0.72 -0.92 1.40 -0.35 -0.66 Post- IIP-32: Non-Assertiveness 9 1.06 0.61 -0.53 0.72 -0.81 1.40 -0.74 -0.58 Pre- IIP-32: Dominance 9 0.94 0.81 -0.26 0.72 -2.19 1.40 -0.36 -1.56 Post- IIP-32: Dominance 9 0.89 0.64 0.35 0.72 -0.52 1.40 0.49 -0.37 Pre- IIP-32: Vindictiveness 9 1.06 0.73 0.74 0.72 1.00 1.40 1.03 0.72 Post- IIP-32: Vindictiveness 9 3.33 3.64 2.95 0.72 8.79 1.40 4.12 6.28 Pre- IIP-32: Self-Sacrifice 9 0.56 0.65 1.19 0.72 0.21 1.40 1.66 0.15 Post- IIP-32: Self-Sacrifice 9 1.44 2.29 1.29 0.72 0.29 1.40 1.79 0.21 Pre- IIP-32: Coldness 9 1.06 0.77 0.66 0.72 0.10 1.40 0.92 0.07 Post- IIP-32: Coldness 9 1.36 0.80 -0.52 0.72 -0.96 1.40 -0.72 -0.69 Pre- IIP-32: Intrusive 9 0.81 0.54 -0.56 0.72 -0.95 1.40 -0.77 -0.68 Post- IIP-32: Intrusive 9 0.81 0.70 0.66 0.72 -0.55 1.40 0.92 -0.39 Pre- IIP-32: Over Accommodation 9 1.31 0.81 0.94 0.72 -0.73 1.40 1.32 -0.52 Post- IIP-32: Over Accommodation 9 0.97 1.00 1.64 0.72 2.98 1.40 2.29 2.13 Pre- IIP-32: Total 9 1.01 0.56 0.85 0.72 0.66 1.40 1.19 0.47 Post- IIP-32: Total 9 1.27 0.81 1.55 0.72 2.34 1.40 2.17 1.67 Valid N (listwise) 9 Note: The cut-off score Z = 1.64 was used to determine whether or not the data is normally distributed. Nonnormally distributed data is marked in bold. 323 The Effectiveness of a Psycho-Educative Group Programme ... Table 6: Normal distribution assessment of the control group data N Mean Std. Deviation Skewness Kurtosis Variables Statistic Statistic Statistic Statistic Std. Err Statistic Std. Err Zs Zk Pre- ECR-R: Total 10 4.01 0.51 0.89 0.69 0.61 1.33 1.29 0.46 Post- ECR-R: Total 10 4.22 0.59 -0.16 0.69 -1.01 1.33 -0.23 -0.76 Pre- ECR-R: Anxious attachment style 10 3.78 0.53 -0.59 0.69 0.46 1.33 -0.86 0.34 Post- ECR-R: Anxious attachment style 10 3.97 0.51 0.78 0.69 1.14 1.33 1.13 0.86 Pre- ECR-R: Avoidant attachment style 10 4.48 0.86 -0.53 0.69 -0.88 1.33 -0.77 -0.66 Post- ECR-R: Avoidant attachment style 10 4.23 0.94 0.16 0.69 -1.33 1.33 0.23 -0.99 Pre- UCLA 10 2.28 0.39 -0.24 0.69 -0.78 1.33 -0.35 -0.58 Post- UCLA 10 2.35 0.38 -0.70 0.69 0.06 1.33 -1.02 0.05 Pre- IIP-32: Social Inhibition 10 1.50 1.05 0.06 0.69 -1.92 1.33 0.08 -1.44 Post- IIP-32: Social Inhibition 10 1.30 0.76 -0.06 0.69 -1.12 1.33 -0.09 -0.84 Pre- IIP-32: Non-Assertiveness 10 1.80 0.75 0.13 0.69 -1.49 1.33 0.19 -1.12 Post- IIP-32: Non-Assertiveness 10 1.40 0.64 -0.04 0.69 -1.01 1.33 -0.06 -0.75 Pre- IIP-32: Dominance 10 1.40 0.91 0.28 0.69 -0.49 1.33 0.41 -0.37 Post- IIP-32: Dominance 10 1.03 0.64 0.21 0.69 0.60 1.33 0.31 0.45 Pre- IIP-32: Vindictiveness 10 1.23 0.46 0.18 0.69 -0.63 1.33 0.26 -0.47 Post- IIP-32: Vindictiveness 10 1.53 0.45 -0.77 0.69 -0.95 1.33 -1.11 -0.72 Pre- IIP-32: Self-Sacrifice 10 0.93 0.73 0.98 0.69 1.46 1.33 1.43 1.09 Post- IIP-32: Self-Sacrifice 10 3.40 3.63 1.74 0.69 3.07 1.33 2.54 2.30 Pre- IIP-32: Coldness 10 1.20 0.85 1.09 0.69 0.88 1.33 1.59 0.66 Post- IIP-32: Coldness 10 0.95 0.64 0.77 0.69 -0.51 1.33 1.11 -0.39 Pre- IIP-32: Intrusive 10 1.43 1.21 1.26 0.69 0.97 1.33 1.83 0.73 324 Nusa Crnkovic Post- IIP-32: Intrusive 10 0.95 0.51 -0.06 0.69 -1.01 1.33 -0.08 -0.76 Pre- IIP-32: Over Accommodation 10 1.03 0.64 0.46 0.69 -0.87 1.33 0.67 -0.65 Post- IIP-32: Over Accommodation 10 0.75 0.46 -0.55 0.69 -1.39 1.33 -0.79 -1.04 Pre- IIP-32: Total 10 1.31 0.55 0.75 0.69 -0.99 1.33 1.09 -0.75 Post- IIP-32: Total 10 1.09 0.35 0.35 0.69 -0.47 1.33 0.51 -0.35 Valid N (listwise) 10 Note: The cut-off score Z = 1.64 was used to determine whether or not the data is normally distributed. Nonnormally distributed data is marked in bold. 325 JAo!!STdVociScri