Exploring the association of staff characteristics with staff perceptions of quality of life of individuals with intellectual disabilities and challenging behaviours

Abstract Background This study aimed to examine the associations between individual staff and staff team characteristics and quality of life of individuals with intellectual disabilities and challenging behaviours. Method With multilevel analyses, we examined educational level, experience, attitudes and behaviours of 240 staff members, in relation to their perception of quality of life of 152 individuals with intellectual disabilities and challenging behaviours they cared for. Results Two individual staff characteristics were related to better quality of life: higher educational and self‐reflection levels. Of the team characteristics, higher educational level, higher self‐efficacy and more friendly behaviour were associated with better quality of life. Unexpectedly, higher staff‐individual ratio was related to lower quality of life. Conclusions Both individual staff and staff team characteristics are associated with quality of life, indicating the need to take staff team characteristics into account when examining quality of life.


| INTRODUCTION
Individuals with intellectual disabilities and challenging behaviours, also defined as 'behaviours of concern' (e.g., Nankervis & Chan, 2021), are at risk of experiencing a lower quality of life than individuals who do not display these behaviours (Simões & Santos, 2017). They more often experience physical injury, social exclusion and lack of autonomy (Allen et al., 2009;Griffith et al., 2013;Matson & Boisjoli, 2009;Sturmey et al., 2005). These are violations of several domains of quality of life as described in the model of Schalock and Verdugo (2002) that distinguishes eight core domains of quality of life: (1) personal development; (2) self-determination; (3) interpersonal relations; (4) social inclusion; (5) rights; (6) emotional well-being; (7) physical well-being; and (8) material wellbeing. According to this model, quality of life is a multidimensional framework that is influenced by the individual themselves (microlevel), as well as their immediate environment (meso-level) and the society they life in (macro-level).
In residential care settings, the immediate environment consists of the direct staff members, who play an important role in the quality of life of the individuals under their care (Claes et al., 2012;Jenaro et al., 2013;Rose, 2011). However, it is unclear how staff characteristics such as age, sex, education level and work experience, but also their attitudes or reactions towards challenging behaviours, are related to the quality of life of the individuals they care for. Knowledge on the relationship between staff team characteristics (e.g., the number of experienced staff members within a team) and quality of life is even more scarce. More insight into the role of staff may provide directions for improving quality of life of individuals with intellectual disabilities and challenging behaviours who live in residential care settings.
The limited quantitative research on the relationship between staff and quality of life of individuals with intellectual disabilities and challenging behaviours has mostly focused on specific domains or indicators of quality of life, such as social inclusion (Bigby & Beadle-Brown, 2018;Claes et al., 2012;Felce et al., 2002a;McConkey & Collins, 2010;Perry & Felce, 2003), emotional well-being (Sexton et al., 2016) or self-determination (Rossow-Kimball & Goodwin, 2009;Stancliffe, 2001;Wehmeyer & Bolding, 2001). Staff characteristics that have been studied in relation to social inclusion are sex, educational level, work experience and staff-individual ratio. Male staff members more often considered stimulating social inclusion of the individuals under their care as part of their job and prioritised it more often than female staff (McConkey & Collins, 2010). Some studies found that higher educational level of staff was related to more social inclusion and self-determination (Mansell et al., 2008;Thomas et al., 1978), whereas in other studies these relationships were not confirmed (Felce et al., 2002a(Felce et al., , 2002b. With regard to work experience, it has been demonstrated that more experienced staff initiated more social activities with the individuals under their care (Mansell et al., 2003), however, they also used more physical restraint (Emerson et al., 2000). Regarding team characteristics, only a very limited number of studies have been conducted, indicating that a higher staff-individual ratio did not enhance quality of life (Beadle-Brown et al., 2016;Felce et al., 2002a). It is important to further investigate team characteristics, since individuals with intellectual disabilities are being cared for by staff teams rather than individual staff members.
Characteristics of one single staff member may have a smaller impact on the quality of life than the characteristics of a complete team. Some qualitative studies have provided valuable suggestions for desirable staff characteristics that may enhance quality of life, based on the views of individuals with intellectual disabilities, staff, and family members (Frounfelker & Bartone, 2020;Petry et al., 2007;Webb et al., 2020;Windley & Chapman, 2010). All respondents considered work experience, sensitivity, self-reflection, friendly behaviour, knowledge, patience, and being caring and empathetic to be important staff characteristics (Petry et al., 2007;Webb et al., 2020;Windley & Chapman, 2010). Hostile staff behaviour, on the other hand, was mentioned as a factor that was expected to decrease quality of life (i.e., lower emotional well-being, less self-determination; Griffith et al., 2013). Furthermore, staff members stated that they need more training on human rights to improve the quality of life of the individuals under their care (Windley & Chapman, 2010), indicating a need for knowledge.
Aside from the few quantitative studies examining the direct relation between staff factors and quality of life, other studies have investigated staff characteristics that may be more indirectly related to quality of life. For example, it was demonstrated that working experience, self-efficacy, attributions of staff about the controllability of challenging behaviours, positive emotions of staff when confronted with challenging behaviours (Willems et al., 2016), and male gender of staff (Willems et al., 2014) were related to more friendly staff behaviour, which has been hypothesised to enhance quality of life (Webb et al., 2020). Moreover, negative staff emotions as a reaction to challenging behaviours were associated with hostile behaviour of staff (Willems et al., 2016), which was mentioned as a possible factor to negatively impact quality of life (Griffith et al., 2013). Self-reflection, higher educational level, proactive thinking and support-seeking behaviour were related to less frequent hostile staff behaviour (Willems et al., 2010(Willems et al., , 2016.
The aim of this study was to expand the body of research on the associations between staff characteristics and quality of life of individuals with intellectual disabilities and challenging behaviours in two ways.
First, we examined the relation of individual staff characteristics (i.e., sex, experience, friendly behaviour, positive and negative emotions to challenging behaviours, attributions about control, self-efficacy, educational level, self-reflection, proactive thinking, assertive control and hostile behaviour) and staff team characteristics (i.e., sex, experience, friendly behaviour, positive and negative emotions to challenging behaviours, attributions about control, self-efficacy, educational level, selfreflection, proactive thinking, assertive control and hostile behaviour) with quality of life of individuals with intellectual disabilities and challenging behaviours in group homes in the Netherlands. We expected male staff, more working experience, friendly behaviour, positive emotions in response to challenging behaviours, belief in external controllability of challenging behaviours, higher self-efficacy, higher educational level, higher self-reflection, more support seeking-behaviour, and more proactive thinking of individual staff or teams to be positively associated with quality of life. Furthermore, we hypothesised that negative emotions in response to challenging behaviours, more hostile behaviours, more assertive control, and more critical expressed emotions of staff or teams to be negatively associated with quality of life.

| Study design
The present study was part of a multi-centre cluster controlled trial, examining the efficacy of a Positive Behaviour Support training for staff in the Netherlands. The trial included four assessments (one pre-test, one mid-test and two post-tests). Data from the pre-test were used for the purpose of this study, resulting in a cross-sectional design.

| Participants
In the Netherlands, each group home for individuals with intellectual disabilities and challenging behaviour has its own specific team of staff providing care for its residents. For our study, we included complete teams of staff and the individuals with intellectual disabilities and challenging behaviours they cared for. The teams were from different organisations in the Netherlands. The individuals the teams cared for and reported on had to meet the following criteria: (1) the individuals were 18 years or older; (2) the individuals had a mild, moderate or severe intellectual disability; (3) the individuals lived in a group home and received 24 h care each day; (4) the individuals interacted with staff at least 2 h a day; (5) the individuals displayed challenging behaviours at least once a week, as assessed by the staff members with the Aberrant Behaviour Checklist (ABC; Aman et al., 1985).

| Procedure
All organisations that were member of the Dutch Association for Care for Individuals with Intellectual disabilities (in Dutch 'Vereniging Gehandicaptenzorg Nederland; VGN') were contacted for participation. Team managers and/or psychologists of interested organisations selected eligible teams and group homes. Meetings in which information was provided about the research project were held for teams that were interested. When staff members consented to participate, similar informative meetings were organised for legal guardians of the individuals with intellectual disabilities. After that, staff contacted the legal representatives for consent for using data about the individual concerned.
For every individual for whom consent was given, a staff member was selected by the psychologist as informant. In the majority of cases, this was the primary staff member (i.e., the staff member responsible for contact with the family and adherence to the treatment plan), alternatively another staff member who knew the individual well. Almost all informants were appointed to one individual with intellectual disabilities and challenging behaviours to conduct the assessments. Three informants completed assessments on multiple (i.e., three) individuals with intellectual disabilities and challenging behaviours.

| Staff self-efficacy: Challenging Behaviour Self-Efficacy Scale
To determine self-efficacy in staff in dealing with challenging behaviours, we used the Challenging Behaviour Self-Efficacy Scale (CBSES; Hastings & Brown, 2002). The CBSES consists of five items covering the following concepts: (1) confidence; (2)

| Emotional reactions of staff: Emotional Reaction to Challenging Behaviour Scale
Positive emotional reactions of staff to challenging behaviours were assessed with the Emotional Reaction to Challenging Behaviour Scale (ERCB; Jones & Hastings, 2003). This 23-item scale comprises a list of positive (e.g., confident, relaxed, cheerful and excited) and negative (e.g., depression, anger, fear and anxiety) emotions that are potentially experienced by caregivers when working with individuals who display challenging behaviours. Each item has to be rated on a 4-point Likert scale ranging from 0 (no, never) to 3 (yes, almost always).

| Challenging behaviours: Aberrant Behaviour Checklist
The primary staff member reported on challenging behaviours of individuals with intellectual disabilities with the Aberrant Behaviour Checklist (ABC; Aman et al., 1985). The ABC is a widely used scale in clinical practice and research, measuring challenging behaviours in individuals with intellectual disabilities. The instrument includes five subscales: Irritability, Lethargy, Stereotypic Behaviour, Hyperactivity/ Noncompliance and Inappropriate Speech. The ABC consists of 58 items with a 4-point Likert scale ranging from 0 (not a problem) to 3 (problem is severe in degree) and has a well-established reliability and validity.

| Statistical analyses
We conducted all analyses twice. First, we examined the association of individual staff characteristics with quality of life of the individual, using the scores of the primary informant. This was in the majority of cases the primary staff member of the individuals with intellectual disabilities and challenging behaviours. Second, we examined the characteristics of staff teams in relation to quality of life of the individual. For the team characteristics, we needed scores that took into account the team as a whole and the individual variation on each measure within the team. We could not use mean team scores, since these only reflect an average of high and low scoring team members, but provide no information on the variation around the mean in teams. Mean scores would thus result in loss of valuable information. Alternatively, in order to better discriminate between teams, we calculated percentages of individual staff members within each team who had high scores on each measure of staff characteristics. These high scores on the measures of characteristics were derived from the explanation of the Likert scale scores of the instrument. We considered the following scores as high scores: '6' or higher on the External Control scale of the CDS (6 meaning 'I agree with this statement'), '5' or higher on the CBSES (5 meaning 'I agree with this statement'), '2' or higher on the ERCBS (2 meaning 'Yes, frequent') and '4' or higher on the SCIBI subscales (4 meaning 'Highly applicable'). For working experience we calculated the mean years of experience of our sample (M = 10.97, rounded up to M = 11) and determined a cut-off score of '12 years' to indicate above average as opposed to mean or below average. With regard to educational level, the percentage of staff members who completed a level 3 education (i.e., higher professional education or master) was used. For the team variable sex we used the percentage of male staff members within a team.
Data were analysed using SPSS software (IBM SPSS Statistics for Windows, version 23.0). We checked the dispersion of the data on every staff and team characteristic. When there was limited dispersion (i.e., less than 15% of the participants scored above the determined cut-off scores), variables were excluded from the subsequent analyses. Additionally, we examined the strength of T A B L E 3 Multilevel analyses of associations between individual staff characteristics and self-determination, social inclusion, emotional wellbeing and total quality of life of individuals with intellectual disabilities and challenging behaviours correlations between variables (i.e., a correlation below 0.3 was considered small, between 0.3 and 0.5 medium and above 0.5 strong; Cohen, 1992), in order to exclude variables in case of strong correlations (i.e., if there was little distinction between variables).
Given the nesting (complete staff teams and individuals with intellectual disabilities and challenging behaviours living together in group homes) and hierarchical structure (staff members within group homes, group homes within organisations) of our data, we conducted multilevel analyses to account for the statistical dependency of the data (Snijders & Bosker, 2012). Multilevel analyses are very suitable for studies which include different levels of aggregation such as our study. Organisation was included as the first level and group home as the second level. First, we conducted the multilevel models for the association between primary staff characteristics and self-determination, social inclusion, emotional well-being and quality of life. Then we repeated the analyses with the team characteristics instead of the primary staff characteristics.
For both multilevel analyses we used a Maximum Likelihood model with an unstructured covariance type. The random effects of the models were organisation and group home. We corrected for the possible confounding effects of the level of intellectual disability and challenging behaviours by adding these to the model as covariates. To distinguish significant associations of the staff characteristics with quality of life (all domains and total), we examined the χ 2 Change (based on À2 log-likelihood and df Change ) with every variable (fixed effect) added to the models. According to the χ 2 distribution, a χ 2 Change larger than 3.84 was considered significant with p < .05 (2-tailed), and Change values larger than 6.63 were considered significant at the level of p < .01 (2-tailed).
The sequence of adding variables to the individual staff models was based on the substantiation in the literature. Variables for which there was quantitative evidence were added first, then variables reported in qualitative studies, and finally variables that were studied in quantitative research but had no or inconclusive quantitative evidence: sex, experience, friendly behaviour, positive emotions to challenging behaviours, hostile behaviour, attributions about external control, self-efficacy, educational level, self-reflection, proactive thinking and assertive control. The sequence of variables of the team models was determined by the size of the χ 2 Change values from the primary staff models (i.e., variables with the largest χ 2 Change first).

| Participant characteristics
In total, 25 group homes were included in the study, containing 240 staff members and 152 individuals with intellectual disabilities who all displayed challenging behaviours. The mean team size was 9.6 staff members (range 4-19). The mean number of individuals with intellectual disabilities and challenging behaviours living in a group home was 6.08 (range 3-9). Tables 1 and 2 present characteristics of staff members and of the individuals with intellectual disabilities and challenging behaviours, respectively.
As is shown in Table 1, the mean scores on the staff variables 'negative emotions', 'support-seeking behaviour' and 'critical expressed emotion' were relatively low and there was limited dispersion towards the higher end of the scale. Therefore, these variables were not included in the multilevel analyses. The dispersion on the team variable 'hostile behaviour' was low with only some clear outliers. Therefore, we also excluded this variable from the team analyses. All other variables were included. T A B L E 4 Multilevel analyses of associations between staff team characteristics and self-determination, social inclusion, emotional well-being, and total quality of life of individuals with intellectual disabilities and challenging behaviours

| Associations between characteristics of individual staff members and quality of life
The multilevel analyses (Table 3) show that education level and selfreflection of staff significantly contribute to the emotional well-being model (χ 2 Change (1) = 4.636, and χ 2 Change (1) = 6.096, respectively). Higher education and higher self-reflection of individual staff members were associated with better emotional well-being of individuals with intellectual disabilities and challenging behaviours (β = 3.499, and β = 0.516, respectively). No other variables contributed significantly to any of the models.

| Associations between characteristics of staff teams and quality of life
Our multi-level analyses, presented in Table 4, demonstrate that education level and the staff-individual ratio contributed significantly to the model of self-determination (χ 2 Change (1) = 5.312 and χ 2 Change (1) = 10.334, respectively); a higher percentage of highly educated staff members within a team was associated with more selfdetermination of individuals with intellectual disabilities and challenging behaviours, and a higher staff-individual ratio was associated with less self-determination (β = 0.037 and β = À5.072, respectively). Further, education level and staff-individual ratio contributed significantly to the social inclusion model (χ 2 Change (1) = 4.615 and χ 2 Change (1) = 8.473, respectively); a higher percentage of highly educated staff members in a team was associated with more social inclusion of individuals with intellectual disabilities and challenging behaviours (β = 0.038). A higher staff-individual ratio was associated with less social inclusion (β = À4.943). Furthermore, staff self-efficacy and friendly behaviour significantly contributed to the emotional well-

| DISCUSSION
In this study, we investigated the association of characteristics of both individual staff and staff teams with quality of life of individuals with intellectual disabilities and challenging behaviours. We found that more self-reflection in individual staff members and teams in which more staff members reported higher levels of friendly behaviour and self-efficacy were associated with better emotional well-being of the individuals under their care. These findings empirically corroborate earlier qualitative studies (Petry et al., 2007;Webb et al., 2020;Windley & Chapman, 2010) in which individuals with intellectual disabilities, parents and staff members perceived friendly staff behaviours and staff self-reflection as important characteristics to increase the quality of life of the individuals concerned. Additionally, our study is the first to find a direct relation of staff self-reflection and friendly staff behaviour with quality of life. This adds to the body of research that found self-reflection of staff was associated with lower hostile behaviour of staff (Willems et al., 2016), which in turn was related to lower quality of life (Griffith et al., 2013). However, to unravel the direction of the associations in order to understand possible causality, more research applying longitudinal designs is necessary. Note: CB, challenging behaviours; *Χ 2 Change is significant at p < .05 (2-tailed), **Χ 2 Change is significant at p < .01 (2-tailed).
As expected, a higher educational level of individual staff members was positively associated with better emotional well-being of the individuals with intellectual disabilities and challenging behaviours, and more staff members within a team with a higher education level was associated with better self-determination, social inclusion and quality of life of the individuals. The findings regarding individual staff confirm results of earlier studies that found that higher education in individual staff members was related with better quality of life of the individuals under their care (Mansell et al., 2008;Thomas et al., 1978).
However, our study indicates that the proportion of higher educated staff members within a team is somewhat differently related to quality of life than higher education in individual staff. The latter was only related to emotional well-being, whereas the first was more broadly

| Strengths and limitations
With this paper, we add to the scarce quantitative studies examining staff and team of staff characteristics that may be associated with quality of life of individuals with intellectual disabilities and challenging behaviours. Never before was total quality of life taken into account, or did studies differentiate between primary staff characteristics and team of staff characteristics. However, our study has limitations. Unfortunately, we could not examine the association between negative staff factors and quality of life of individuals with intellectual disabilities and challenging behaviours. Our data on negative staff emotions in response to challenging behaviours, staff critical expressed emotions and hostile staff behaviour showed overall low scores and little dispersion and could therefore not be included in the analyses. Although the scores on these variables were comparable to other studies that used the same instruments (Jones & Hastings, 2003;Willems et al., 2014Willems et al., , 2016, it is questionable if these negative factors did indeed not occur in our sample or if the staff members answered socially desirable, which is a known concern (Lambrechts et al., 2010). Some staff members may experience a mismatch between their initial response to challenging behaviour and dependent variables (i.e., quality of life). Therefore, we recommend future studies to take more proxies into account for different perspectives and greater reliability. Another limitation is that we did not include all potentially important staff and team of staff characteristics in our study. For example, it would be interesting to examine the effect of (PBS) staff training, organisational culture, staff turnover, job satisfaction of staff, heterogeneity within staff teams, the relationship of individuals with intellectual disabilities with their family members or the relationships between staff members into account (Bigby & Beadle-Brown, 2018). Furthermore, it is important to improve our insight into the associations of various individual characteristics such as the presence of comorbid disorders, physical well-being, and housing situation with the quality of life of individuals with intellectual disabilities and challenging behaviours. Future studies should include these factors when examining relations between characteristics of (teams of) staff individual characteristics with quality of life of individuals with intellectual disabilities and challenging behaviours.

| Clinical implications
The results of our study provide insights with relevance for clinical practice. First, we would like to highlight the need to educate staff members in how their attitude may be related to the quality of life

| CONCLUSIONS
Our study was the first to take both individual staff and staff team characteristics into account to examine their relation with quality of life of individuals with intellectual disabilities and challenging behaviours. We found that higher educational level and self-reflection of individual staff members were related to better emotional well-being, and, in staff teams, that higher educational level, friendly behaviour and self-efficacy were associated with higher quality of life (i.e., selfdetermination, social inclusion, emotional well-being, and total quality of life). The associations between higher staff-individual ratios and lower self-determination and social inclusion were unexpected. Remarkably