Attributions of aggressive behaviour in people with mild intellectual disabilities to borderline intellectual functioning in a secure forensic setting

ABSTRACT

Aggressive behaviour is often displayed by people with intellectual disabilities (ID) in forensic healthcare settings. Research on the causal beliefs (i.e. attributions) of aggressive behaviour are commonly studied from the perspective of support staff. As aggressive behaviour is mostly a product of interaction between the person showing it and their environment, it is valuable to include the perspective of people with ID as well. Four group interviews, consisting of a total of 20 people with mild ID or borderline intellectual functioning and forensic and/or psychiatric problems, were held to explore incidents of aggressive behaviour. The attributions were analysed using the Leeds Attributional Coding System. Clients almost equally distributed the causes of aggressive behaviour to themselves (intrapersonal domain; 48.0%) and to other persons (interpersonal domain; 45.7%). There is a distinction related to the attributions given between the client as agent (intrapersonal domain), being uncontrollable (72.1%) and global (68.9%), versus other persons as agent (interpersonal domain), being controllable (86.2%) and specific (56.9%). This analysis of attributions regarding aggressive behaviour given by clients resulted in information on causal beliefs of aggressive behaviour from the perspective of clients. Incorporating their views will possibly increase involvement and commitment in support and treatment.

KEYWORDS:

• Attributions
• aggressive behaviour
• mild intellectual disabilities
• borderline intellectual functioning
• forensic setting

Background

Aggressive behaviour is a well-recognised issue in forensic mental health settings (Alexander et al., 2010; Bowers et al., 2011). Prevalence rates of aggressive behaviour within these settings are especially high for people with mild intellectual disability (MID) or borderline intellectual functioning (BIF) compared to people without ID (Chester et al., 2018; O’Shea et al., 2015). For example, in their study in a secure mental health hospital in the UK, O’Shea et al. (2015) reported rates of aggressive behaviour of people with mild ID of 83.5% compared to 61.3% for people without ID. People with MID and BIF account for a great number of aggressive incidents within forensic mental health settings, even though they do not form the largest forensic population (Baldry et al., 2013; Dickens et al., 2013).

Aggressive behaviour can have negative consequences for both the aggressors themselves (e.g. physical injuries, decreased quality of care, increased restrictions, and higher medication use; Arnetz & Arnetz, 2001; Bowers et al., 2011; Deb et al., 2015; Tsiouris, 2010; Van den Bogaard et al., 2018) as well as their environment supporting them (e.g. negative emotions, stress and burnout, physical injuries, and high costs of care; Bowers et al., 2011; Erdos & Hughes, 2001; Griffith & Hastings, 2014; Mills & Rose, 2011; Singh et al., 2008). As these consequences are detrimental, a significant amount of research has already been conducted in various settings, including the secure forensic setting, to gain more insight into the prevention and management of aggressive behaviour. So far, most of this research on aggressive behaviour focused on the perspective of the environment, by concentrating on the views of support staff, families, and other proxies (e.g. Bowers et al., 2011; Deveau & McGill, 2014; Hassiotis et al., 2018; Jacobs et al., 2016). Consequently, the views of people with MID and BIF themselves related to aggressive behaviour are understudied. Incorporating the views of people with MID or BIF in care and support is, however, important for three main reasons. First, according to clients and support staff, being involved and engaged in services and therapies as a client is a precursor for progress in these areas (Morrissey et al., 2017). Second, support staff and people in psychiatric care have different opinions regarding the causes of aggressive behaviour and therefore both views should be studied (Duxbury & Whittington, 2005) to develop optimal interventions. For example, Duxbury and Whittington (2005) found that, clients attributed aggressive behaviour primarily to environmental conditions and poor communication, whereas nurses attributed aggressive behaviour mainly to the clients’ mental illnesses and in-client environment. Third, based on their review, Bowers et al. (2011) showed that the antecedent ‘no clear cause’ was one of the themes with the highest proportion precipitating aggressive incidents (i.e. 32%). Hence, asking clients about these causes might be helpful. Although the three studies do not specifically report on people with MID or BIF in a secure forensic setting, they do underline the importance of incorporating the clients’ perspective regarding causes of aggressive behaviour.

In order to identify the available studies focusing on the causal beliefs (i.e. attributions) of challenging behaviour according to people with ID themselves, Van den Bogaard et al. (2019) conducted a systematic review. Two of the included studies (N = 10) focused on the experiences of people with ID (level of ID was not specified) regarding aggressive behaviour. That is, Clarkson et al. (2009) asked participants about their perceptions and experiences regarding direct support staff. In their study, the only reported attribution of aggressive behaviour by people with ID was dishonesty of support staff. In addition, Fish and Culshaw (2005) explored the explanations of people with ID regarding aggressive incidents and the consequences of physical interventions, revealing the atmosphere on the ward, the locked environment, and support staff (e.g. support staff did not listen to the client) as causes of aggressive behaviour.

In sum, although the prevalence of aggressive behaviour in people with MID or BIF in forensic settings is high and they do account for a great part of the aggressive incidents, the attributions of people with MID or BIF displaying aggressive behaviour themselves in these settings are not yet explored. Therefore, in this study, we explored the attributions regarding aggressive behaviour of people with MID or BIF within a secure forensic psychiatric setting. More insight into their attributions possibly helps them to become more motivated for support and treatment (Morrissey et al., 2017). Besides, as people with MID or BIF are able to tell how they felt before the aggressive behaviour, it might help support staff to become more sensitive towards these emotional states and to intervene more proactively (Bowers et al., 2011).

Methods

Participants and setting

In total, 20 individuals with MID (IQ 50–70) or BIF (IQ 70–85) participated in this study. Participants resided on average for 20 weeks (range: 3–49 weeks) in a secure forensic psychiatric hospital in the Southern part of the Netherlands. Within this setting, they received treatment for their forensic and psychiatric problems (e.g. substance-related disorder or psychosis). The participants (19 men) had a mean age of 35.2 years (SD = 8.9; range: 20–53 years). The level of intellectual functioning was determined based on data from psychometric sound IQ tests as described in the participants’ files. The mean IQ on file was 63 (SD = 7.9; range: 50–83, for 2 participants no specific IQ scores were available, but the psychologist of both clients reported IQ-scores between 50 and 85 based on their clinical expertise); 16 were diagnosed with MID, 4 participants with BIF. In addition to their forensic problems, 95% of the participants were diagnosed with one or more comorbid psychiatric diagnoses using the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM, American Psychiatric Association, 2013), of which substance-related disorders (n = 18), schizophrenia spectrum and other psychotic disorders (n= 4), and personality disorders (n = 4) were the most common diagnoses.

Procedure

After ethical approval from the Ethics Review Board of Tilburg University (EC 2014.21) and the participating healthcare organization, the researchers informed all support staff of the forensic psychiatric hospital about the aim and content of the study during a team meeting. The purpose of informing them was to alleviate any potential concerns regarding the invasiveness of the research for the clients, to involve them in starting this research and to support clients in case they would have any questions regarding the study. Next, in two information sessions, the researchers informed all potential participants (i.e. all persons with MID or BIF residing at the secure forensic psychiatric hospital) about the goal, content, and procedure of the study, using visual aid. That is, the researcher explained that, after the information session, clients would be invited to participate in a group interview by one of their support staff members. It was explained that, during the group interview, they would be asked to provide descriptions of what happened before, during, and after incidents of aggressive behaviours on the ward (in general) in which they were involved or which they had witnessed. Moreover, during the information sessions, it was explained, both verbally and written on the information sheet and informed consent form, that participants would not be named in reports. Two support staff members were present at each information session as they knew the clients well and answered their questions and clarified their indistinctness.

After written consent to participate, participants were informed about the time and location of the group interview. In total, four group interviews were executed, one on each ward of the secure forensic psychiatric hospital. In order to make the clients feel comfortable, a psychologist with experience in ID research and an experienced staff member familiar with the target population and clients executed the group interview. The researcher gave an extensive briefing and checked the interviews to confirm open questioning, without influencing the participants. Using familiar staff to execute the group interview is considered to help the participants to feel comfortable and helps to elicit most information (Norman & Parker, 1990). Besides, the advantage of a group interview over an individual interview is that group interviews are less confrontational, not directly related to one’s own incident, and it might feel safer to talk in a group (e.g. Kitzinger, 1994). In addition, a video camera instead of a voice recorder was used after consultation with the participants themselves; a voice recorder reminded them of negative events in the past (e.g. conviction at a police station) and the position of the camera was focused on the interviewers instead of the clients. Confidentiality and anonymity of the information was emphasised; it was explained that the group interviews would be transformed into text, without disclosing the participant’s identities.

The group interviews took place in a familiar room in the forensic psychiatric hospital. To make the clients feel as comfortable as possible and to avoid clients to see this conversation as a regular talk with the psychologist, the group interview took place in the recreational room of the forensic psychiatric hospital. Refreshments were provided, and participants received a small financial reward (€10) for their participation. On average, the group interview had a duration of 32 minutes (range: 28–44 minutes). Two group interviews consisted of five participants, one consisted of four and one of six participants.

Instrument

To get a full description of incidents of aggressive behaviour among people with MID and BIF with additional forensic and psychiatric problems in a secure forensic setting, we used a semi-structured interview schedule based on van den Bogaard et al. (2020) who focussed on the attributions of support staff working with people with MID or BIF. Using open-ended questions to elicit non-biased responses of the participants, they were asked to give a description of what happens when a person (the client themselves or other clients on the ward) became aggressive. In case the answer was unclear or incomplete, the interviewers asked clarifying questions related to (1) the antecedent (i.e. the triggers of aggressive behaviour, such as closeness or denial of request), (2) the aggressive behaviour (i.e. the specific topography of the behaviour, such as verbal or physical aggression), (3) the target of the behaviour (i.e. the person or materials to whom the behaviour was aimed at, such as other clients or support staff), (4) the consequences of the behaviour (i.e. consequences of the behaviour for the victim, such as feeling threatened), (5) the measures that were taken to stop the aggressive behaviour (i.e. what was done to stop the behaviour, by the person him/herself or the environment, such as going to own room or holding), and (6) feelings of the aggressor or victim related to the behaviour (i.e. what did the persons involved felt before, during or after the incident?). Adaptations to the interview schedule of van den Bogaard et al. (2020) were made to ensure that the questions were comprehensible for people with MID or BIF (e.g. simple and clear language and visualization of the information).

Analysis

The analyses were carried out in two steps. First, a qualitative content analysis was conducted to explore the content of the attributions of participants. The content analysis was executed by the first author by deductively coding the attributions based on the code list derived from the review of van den Bogaard et al. (2019), originating from the models of causation for aggressive behaviour (Duxbury, 2002; Nijman et al., 1999). The complete coding was checked by a second ID researcher, and disagreements were discussed until consensus was reached. Second, in addition to the content analysis, descriptive statistics were used for identification of the agent and exploration of the scores on the five attributional dimensions of Leeds Attributions Coding System (LACS).

The Leeds Attributional Coding System (LACS; Stratton et al., 1988) was used to analyse the data of the group interviews. The LACS is a method developed to explore attributions made by one person, based on individual or group interviews (Munton et al., 1999). The research method is designed to explore transcripts in six steps. That is, each group interview was recorded and transcribed verbatim, to indicate the source of the attribution (step 1). Next, the transcripts of the group interviews were independently screened by two researchers (first author and an experienced ID researcher) to identify which text contains attributions of aggressive behaviour (step 2). Attributions are ‘expressions of the way a person thinks about the relationship between a cause and an outcome’ (Munton et al., 1999, p. 6). In this study, attributions that were formulated by the interviewer and only confirmed (and not elaborated on) by the respondents were not taken into consideration for the analysis. The extractions were compared and differences were discussed until consensus was reached. Following this, the first author separated the cause (e.g. I (client) did not like what support staff said to me (= cause)) and outcome (e.g. that is why I (client) called her (support staff) names (= outcome)) in the attributions (step 3) and identified the speaker (i.e. the person providing the attribution), agent (i.e. the person or situation mentioned in the cause of the attribution), and target (i.e. the person or situation mentioned in the outcome of the attribution) (step 4). Subsequently, the attributions were coded by the first author on five causal dimensions (see Appendix A; step 5). To calculate the Percentage of Agreement Index (Suen & Ary, 1989), a second ID researcher coded 25% of the group interviews; the percentage of agreement was 91.4%. Last, the codes were analysed to conduct descriptive and comparative analyses (step 6).

Results

In this section, the content of the attributions will be described first, followed by an overview of the scores of the attributions on the five attributional dimensions of the LACS per type of agent (Stratton et al., 1988).

Content of the attributions

In total, in the four group interviews, participants provided 127 attributions about aggressive behaviour, differentiating between attributions on the interpersonal, environmental, and intrapersonal level.

Regarding the interpersonal attributions, participants mentioned triggers related to support staff, other clients and other persons not specified as causing them to display aggressive behaviour. Although participants did not specify other persons in these attributions, based on the description of the incidents, it seemed that these persons were closely related to the participants, such as other clients or support staff. Participants mentioned the reactions of support staff and other clients, but also the lack of reactions by support staff as triggers for aggressive behaviour.

… and they [support staff] sometimes treat you like a child. Because, I’m 54 years old, you are not going to tell me: ‘Go to your room!’. You can say it differently: Would you like to go to your room?

Or

If you [support staff] or somebody else pretends that he wants to help me, trying to ruin my life, then we go together.

Besides, participants mentioned various negative feelings as a consequence of the behaviour of other people (e.g. support staff or other clients) which causes them to show aggressive behaviour, such as being hurt, belittled and feeling abandoned. As one client tells why he became aggressive: ‘If you live in the same area and people belittle you, behind your back’. Other feelings that were mentioned by the participants causing aggressive behaviour were annoyance, irritations and feeling one have to defend oneself towards other people. As one clients tells why he became aggressive: ‘Because they (other clients) irritate me’.

Environmental factors contributing to aggressive behaviour were related to the physical context, feeling socially excluded and situational factors. That is, participants mentioned being at the ward and not with family, the atmosphere on the ward, and the untidiness of other clients as situational factors causing aggressive behaviour:

… you have to deal with the situation that you’re locked in here and you’re not with your own family

According to one client, the reason for him to become aggressive was a sequence of different factors: ‘It a sequence of the situation at home and dealing with the fact that you are here and not with your family. Although you can have some contact, they are not around at this moment’.

Last, regarding the intrapersonal attributions causing aggressive behaviour, participants mentioned nine different types of reasons: (1) psychiatric disorder (e.g. having ADHD). For example:

‘One person has borderline, others have ADHD ADD, or autism spectrum disorder and that kind of stuff. Because of that, everybody reacts differently’.

(2) Medical reason (e.g. epilepsy), (3) mental health (e.g. brooding), (4) emotions and feelings (e.g. anger):

(5) Coping with these emotions and feelings (e.g. not being able to talk about feelings), (6) life history (e.g. things that happened in the past), (7) status (e.g. wanting to get things done), (8) positive reward (e.g. feeling calm afterwards), and (9) other (e.g. becoming aggressive by talking about incidents of aggressive behaviour).

Attributional dimensions per type of agent

The scores on the attributional dimensions per type of agent are displayed in Table 1. Intrapersonal attributions are primarily scored as unstable, global, internal, personal, and uncontrollable, suggesting that the causes of aggressive behaviour are not likely to influence future outcomes (unstable; e.g. ‘Well yes, if I (client) am not able to explain it with words, by talking, that’s when I really break out’). It also suggests that the causes of aggressive behaviour are likely to influence several different outcomes as well (global). For example, the following cause mentioned by a client, is likely to have many other outcomes in his life, besides becoming aggressive: ‘There are clients who have a borderline personality disorder’. He also might have difficulties is maintaining social contact with friends. The causes further originate from within the person (internal; e.g. ‘one person has ADHD, everybody reacts differently’), and tell something unique about the person (personal; e.g. ‘I can get things done, the way I want to’).

Table 1. Scores on five attributional dimensions per type of agent

	Agent
	Intrapersonal		Interpersonal		Environmental		Total
	n	(%)	n	(%)	n	(%)	n	(%)
Stable	13	(21.3)	12	(20.7)	3	(37.5)	28	(22.0)
Unstable	48	(78.7)	46	(79.3)	5	(62.5)	99	(78.0)
Global	42	(68.9)	24	(41.4)	5	(62.5)	71	(55.9)
Specific	18	(29.5)	33	(56.9)	3	(37.5)	54	(42.5)
Not coded	1	(1.6)	1	(1.7)	0	(0.0)	2	(1.6)
Internal	56	(91.8)	54	(93.1)	0	(0.0)	110	(86.6)
External	5	(8.2)	2	(3.4)	0	(0.0)	7	(5.5)
Not coded	0	(0.0)	2	(3.4)	8	(100.0)	10	(7.9)
Personal	52	(85.2)	52	(89.7)	0	(0.0)	104	(81.9)
Universal	9	(14.8)	3	(5.2)	0	(0.0)	12	(9.4)
Not coded	0	(0.0)	3	(5.2)	8	(100.0)	11	(8.7)
Controllable	17	(27.9)	50	(86.2)	0	(0.0)	67	(52.8)
Uncontrollable	44	(72.1)	5	(8.6)	0	(0.0)	49	(38.6)
Not coded	0	(0.0)	3	(5.2)	8	(100.0)	11	(8.7)
Total	61	(48.0)	58	(45.7)	8	(6.3)	127	(100.0)

In contrast, interpersonal attributions (i.e. support staff, other clients, or others (not specified) were the agent) were scored as unstable, internal, and personal, mostly specific and controllable. This suggests that, when other people were the agent, causes of aggressive behaviours were not likely to influence many other outcomes, except for becoming aggressive (specific; e.g. ‘Others (not specified) want to have power over you’) and that the agent was able to influence the outcome (e.g. controllable; ‘Support staff do not react’).

Discussion

Aggressive behaviour displayed by people with MID or BIF is a common phenomenon in forensic mental health settings, though studying causal beliefs of this behaviour from the perspective of people with MID or BIF themselves is rare. However, their perspective is important as it can provide a rich inside in their explanations of their behaviour and, as a result, help to develop more effective interventions. Therefore, this study explored the attributions regarding aggressive behaviour of people with MID and BIF and additional forensic and psychiatric problems residing in a secure forensic psychiatric hospital. In order to do so, four group interviews were held with 20 clients and transcribed verbatim to extract attributions regarding aggressive behaviour. Next, the content of the attributions provided by the participants were analysed. Subsequent, we examined which persons and what situations were mentioned in the cause of the attribution (agent) and we explored the scores on the five attributional dimensions (stable/unstable, global/specific, internal/external, personal/universal, and controllable/uncontrollable) according to the LACS (Stratton et al., 1988) to indicate the characteristics of the attributions.

Based on the content analysis, it can be concluded that participants attributed aggressive behaviour to a broad range of interpersonal, environmental, and intrapersonal factors. As such, participants emphasised their own role as well as the role of their environment (support staff, other clients, and other persons (not specified)) in causing aggressive behaviour. More specifically, in about half of the cases, participants mentioned characteristics and behaviours of themselves causing CB, such as having a personality disorder or feeling distressed. In the other half of the attributions, participants indicated characteristics other clients and support staff as causing aggressive behaviour. They indicated various feelings that were evoked in them by behaviours of other people (e.g. getting frustrated because support staff treat the client like a child), which eventually led to aggressive behaviour. Last, in some cases, they identified situations as causing aggressive behaviour. These findings are comparable to recent findings exploring the motives of patient without an ID in a forensic psychiatric care setting regarding aggressive behaviour (Lewis & Ireland, 2019). Also in this study, patients mostly reported intrapersonal and interpersonal causes for aggressive behaviour.

Besides comparing the views of people with and without ID, it is also informative to compare the views of people with ID and support staff. Already many studies have examined the attributions of support staff related to challenging behaviour and more specific aggressive behaviour (e.g. Noone et al., 2006; van den Bogaard et al., 2019). In the study of van den Bogaard and colleagues (2020) it became clear that support staff mentioned intrapersonal client factors in 70% of the attributions related to aggressive behaviour. Support staff attributed interpersonal factors as unstable, specific, internal, universal and controllable and intrapersonal factors as unstable, global, internal, personal and uncontrollable. Although these results of support staff are different compared to clients (higher percentage of intrapersonal attributions and interpersonal attributions labelled as personal), support staff did not work in a forensic psychiatric hospital and as such this can impact the results given. Besides, the results of other studies that compared the views of support staff and clients without an ID (Lamanna et al., 2016), indicated mixed results as well. In future research and clinical practice, it would be helpful to get more information of both perspectives as this can be a starting point to initiate a conversation in which clients and support staff can learn to understand each other and the reasons for their behaviour. This can be helpful in fostering a greater understanding of the needs, behaviours and motivations of a client and in deciding together the most appropriate interventions for these complex behaviours.

Next, the results provide support staff more insight into the emotional state of a client prior to an incident, and, as such, provide a greater psychological understanding of the client. Asking people with MID and BIF and forensic and psychiatric problems about their perspectives on the causes of aggressive behaviour – and probably also for other forms of challenging behaviour – might also help support staff to become aware of their contribution in triggering and maintaining aggression, and to observe not only behavioural but also emotional cues (Bowers et al., 2011).

Moreover, getting insights in the emotions and feelings experienced by clients precipitating aggressive behaviour might also be helpful for the clients themselves. As people with ID have fewer psychological resources to cope with stressful events (Van den Hout et al., 2000), it would be very helpful to get more insights into their own emotional experience before and during a negative event (e.g. CB). Several interventions, such as Cognitive Behaviour Therapy (CBT) or self-management training, can be used effectively to develop cognitive skills and self-control techniques in people with ID to manage their mental health and emotional problems (Beail, 2003; Taylor & Novaco, 2005; Willner, 2005). Training individuals to manage their own emotions and feelings makes them less dependent on their environment (Taylor & Novaco, 2005) and possibly more motivated for change (Morrissey et al., 2017). Although these studies did not focus specifically on aggressive behaviour in our population, it would be valuable to test if these training courses are helpful for aggressive behaviour in people with MID or BIF with forensic and psychiatric problems.

Clients also mentioned interpersonal attributions related to aggressive behaviour. This confirms results from earlier studies in which the role of support staff in triggering and maintaining aggressive behaviour, but also in the prevention and management of aggressive behaviour is acknowledged by support staff (e.g. Nijman & À Campo, 2002; Van den Bogaard et al., 2018) and clients (e.g. Griffith et al., 2013; Jones & Stenfert Kroese, 2007). For example, support staff actions can trigger aggressive behaviour, but also their reactions (e.g. restrictive interventions aimed at preventing dangerous situations), could paradoxically trigger the repeated occurrence of aggressive behaviour (Griffith et al., 2013; Nijman et al., 1999). Clients, in a recent review of van den Bogaard and colleagues (2019) report that people with ID have identified the attitudes and reactions of support staff (or their lack of reaction) as triggers for aggressive behaviour.

Given the potentially major impact of CB on support staff, support staff might have difficulties understanding and responding to CB (Whittington & Burns, 2005). Training and coaching support staff to understand their own attributions and related behaviour – and, more specifically, to understand their influence on the existence and maintenance of CB – is therefore likely to help improve the reactions of support staff to CB. As support staff have the task on the one hand to support and take care of the client (Eager et al., 2007), and on the other hand to deliver behavioural interventions as well to prevent and manage CB (Allen, 1999), they should be equipped with the right knowledge, skills and attitudes to deal with behaviour that challenges (NICE Guidelines; National Institute for Health and Care Excellence (NICE), 2015). Knowledge for example about the behavioural principles and functionality of CB, skills to examine the functionality of CB and to deliver behavioural interventions based on this functionality of CB are important (National Institute for Health and Care Excellence (NICE), 2015). And last, a more empowering, instead of controlling attitude towards clients with ID and CB is developed (Randell et al., 2017).

At this stage, many different training programmes are developed for support staff of people with ID and CB (e.g. Willems et al., 2016; Zijlmans et al., 2011). These training programmes do focus on the behaviour of support staff and help them to attune their reactions to the behaviour of the client. Moreover, if support staff have more insight in their own emotional reactions, this will help them to react in a more favorable way towards clients with ID and CB (Zijlmans et al., 2015). By supervising and coaching (both in training and on the job), support staff becomes more aware of the subjective impact of CB on themselves (e.g. their negative emotions), which will help them, through self-reflection, to react in a way that will reduce CB (e.g. less controlling behaviour; Willems et al., 2016).

Next, the attributions were analysed by scoring them on five attributional dimensions. Related to intrapersonal attributions, clients attributed most causes of aggressive behaviour as unstable, internal, personal, global, and uncontrollable. Interpersonal attributions (e.g. support staff, other clients or persons as agent) were also scored as unstable, internal and personal, but clients also attributed the behaviour as specific and controllable. This suggests that participants attribute causes of aggressive behaviour to aspects which are more under the control of their environment and which do not influence many other outcomes. Getting to know more about the attributional dimensions of aggressive behaviour of clients might be helpful to connect with their experiences and hence to attune treatment to their wishes and needs (Morrissey et al., 2017). This can be done within an autonomy supportive environment, where support staff minimizes control while supporting clients’ initiatives and accepting their perspective (Williams et al., 2006). A person with ID who feels more autonomous within his or her environment, is more involved and engaged in one’s own treatment, which can eventually lead to better treatment outcomes (Frielink et al., 2018; Morrissey et al., 2017). For example, if someone attributes aggressive behaviour mostly as external to himself, that person might not take responsibility for his behaviour, and hence, this might imply a lack of progress in treatment due to not taking responsibility for his behaviour. Or, if a person attributes aggressive behaviour as stable (i.e. not likely to change in near future), this can result in a passive attitude as the person does not experience possibilities to change the situation. So, knowing the way a person attributes his behaviour gives the person himself more insight into his own behaviour and the relation with their environment, but also provides their environment information about the reason this person might act in a certain manner. Both can be helpful in preventing aggressive behaviour. It would therefore be recommendable to examine both the attributions of clients and staff members to direct treatment and support. Finally, due to the small number of attributions, it was impossible to analyse attributional styles (i.e. patterns on the five attributional dimensions) of people with MID or BIF with forensic and psychiatric problems. Focusing on attributional styles and patterns in future research can give insights in adaptive and maladaptive ways of thinking and relating behaviour, cognitions and interrelated emotions (Dix & Reinhold, 1991).

Although this study relies on the use of qualitative data based on real incidents and gives rich information of the experiences of people with MID or BIF and forensic and psychiatric problems regarding aggressive behaviour, the results of this study should be interpreted in the light of several limitations. First, the Leeds Attributional Coding System relies on verbal data extracted by transcriptions of respondents, retrospectively asked. As we do know that people with ID can have difficulties in communication and emotional inspection, we have made some arrangements to overcome these difficulties. During data collection, we used visual support to help the clients understand the questions and we adapted the interview guide into easy accessible language. During the analysis phase, we only analysed attributions that were given or elaborated on by the participants, and we did not use attributions of the interviewer, which were only confirmed by the participants. Next, the sample in this study is predominantly male. Therefore, findings may not generalize to women with ID in forensic psychiatric settings. Last, data collection took place in one secure forensic psychiatric hospital in the Netherlands and relies on retrospective information retrieved from the participants. In future research, it would be recommendable to extract more data derived from real incidents in different settings and related to different behaviours (e.g. self-injurious behaviour) and different types of behaviours (e.g. verbal versus physical aggression) as well as linking this information to individual characteristics, behavioural and treatment outcomes. Moreover, it would be recommendable for future research to interview both people with MID and BIF and support staff about the same incident. This will give more insight into each other’s perspective and may provide opportunities to develop treatment interventions that connect both the needs of people with MID and BIF and forensic and psychiatric problems with possibilities of support staff in preventing or reducing aggressive behaviour.

In sum, this study showed that people with MID and BIF and forensic and psychiatric problems are capable of describing causes of incidents of aggressive behaviour in detail, making it possible to extract attributions. The diversity in the answers of participants in the current study regarding the contribution of the clients themselves, support staff, other clients and situations in triggering aggressive behaviour suggest that people with MID and BIF are sensitive to the internal and environmental contributing factors of aggressive behaviour. Although the incorporation of the views of people with MID and BIF in forensic mental healthcare is still developing, this study shows the potential of incorporating the views of this population and also to embed this information in a broader context of support and treatment. Understanding behaviour as a product of interaction between the person showing the behaviour and their environment (Banks et al., 2007) starts with recognizing that the views of both professionals and clients are valuable and, thus, in order to manage this behaviour, people with MID and BIF in forensic mental healthcare need to be asked about their opinion as well.

Acknowledgments

We would like to thank all participants, their support staff and the interviewers for their contribution to this study. A special thanks to Annemarie Kroon for her help related to the preparation and collection of data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Quality of Forensic Care [Kwaliteit Forensische Zorg (KFZ)] under Grant 2016-57.

Appendix A

Definition of causal dimensions used in the Leeds Attributional Coding System.

Table

Causal dimension	Definition of causal dimension
Stable/unstable	Stable causes are causes that are likely to continue to influence outcomes in the future
Global/specific	Global causes are those cause that are likely to have a significant impact on several different outcomes
Internal/external	Internal causes are those believed to originate from within the person being coded
Personal/universal	Personal causes contain information concerning something unique or idiosyncratic about the person being coded
Controllable/uncontrollable	Controllable causes are causes in which the person thinks he could have influenced the outcome