Published for the British Institute of Learning Disabilities

Journal of Applied Research in Intellectual Disabilities 2015, 28, 589–593

BRIEF REPORT

Exploring Carers’ Judgements of Responsibility and Control in Response to the Challenging Behaviour of People with Intellectual Disabilities Sophie Williams*, Dave Dagnan†, Jacqui Rodgers‡ and Mark Freeston‡ *Northumberland Head Injury Service, Northgate Hospital, Morpeth, Northumberland, UK; †Cumbria Partnership NHS Foundation Trust, Community Learning Disability Services, Lakeland Business Centre, Workington, UK; ‡Clinical Psychology, University of Newcastle, Newcastle upon Tyne, UK

Accepted for publication 14 January 2015

Background This study examines Weiner’s recent cognitive emotional model which makes a distinction between judgements of control and responsibility and emphasizes the moderation of control by ‘mitigating’ factors. Method In response to four vignettes describing two conditions of control (high or low) and mitigating factors (present or absent), questionnaires rating judgements of responsibility and emotional responses (anger and sympathy) were completed by 52 care staff. Results Analysis of the data for sympathy demonstrated that attributions of control were moderated by

Introduction Understanding carer responses to challenging behaviour is important in developing high-quality services to people with intellectual disabilities and there is a growing body of work in this area. Weiner’s attribution model of motivation (1985) has attracted interest in explaining how cognitive variables can play a part in carer behaviour (e.g. Dagnan et al. 1998). More recently, Weiner (1995) has developed the model to focus on judgements of responsibility. In the revised model, Weiner differentiates between control (an attribution) and responsibility (a moral judgement). Weiner suggests that when observing an event or behaviour, we first attribute a level of control and then, if the attribution is of a high degree of control, we may seek to mitigate this © 2015 John Wiley & Sons Ltd

communication ability and that the effect of control upon sympathy was mediated by the judgement of responsibility. Conclusions The data offer tentative support Weiner’s account of the mitigation of control attributions in making responsibility judgements and their subsequent effects on emotional responses. Implications for research and clinical work are discussed. Keywords: attribution, carers, control, responsibility

challenging

behaviour,

attribution through understanding the reasons for the behaviour. An example is a person who intentionally crashes their car (a controllable cause for the crash) but who may not be held significantly responsible because the act was carried out to avoid hitting a child who ran on the road (a mitigating circumstance; Weiner 1995). The presence or absence of a mitigating circumstance may therefore significantly effect the observer’s judgement of responsibility and subsequent emotional and behavioural response (in the case of carers, the effort expended in helping and possibly the type of helping offered). Few studies have included responsibility as a variable in studies of carers of people with intellectual disability (e.g. Dagnan & Cairns 2005). However, some studies have effectively included mitigating factors in their 10.1111/jar.12146

590 Journal of Applied Research in Intellectual Disabilities

design. For example, Stanley & Standen (2000) found that there was less anger reported for aggressive behaviour (which tends to be seen as a ‘controllable’ behaviour) when people were described as more disabled than when they were described as less disabled; however, these studies do not distinguish between control and responsibility. This study investigates Weiner’s (1995) model when applied to carers working with people with challenging behaviour. The specific predictions explored are that the effect of carers’ attributions of control on responsibility and emotions will be moderated by the presence or absence of a mitigating factor and that the moderation effect will particularly evident in conditions of high control. Whilst the moderated effect of control will have some effect on the levels of emotions, this effect will be predominantly mediated by responsibility.

Method Participants Eighty-four carers in supporting living and residential care roles were invited to participate and 52 (62%) returned questionnaires of which 50 (60%) had sufficient data to be fully included in the analysis. The group consisted of eight men (16%) and 42 women (84%), had a mean age of 41.7 years (SD = 11.4) and a mean length of experience in working with people with intellectual disabilities and challenging behaviour of 8.9 years (SD = 6.3).

Design The study used a repeated-measures design. Four distinct vignettes describing instances of challenging behaviour were created describing two conditions of control (high or low) and mitigating factors (present or absent). The two by two vignette design allowed the direct testing of the moderating effect of mitigating factors on responsibility and emotion using analysis of variance (ANOVA).

Development of vignettes and measurement of variables All vignette materials were piloted over two phases, and the results from each pilot were then graphed and examined visually to check whether they indicated that manipulation of control and mitigation were achieved across the four vignettes.

The final four vignettes described ‘Mark, a 35 year old man living in a residential home for people with learning disabilities’ and represented combinations of the two factors of interest (high and low personal control, the presence and absence of mitigating factor). Each contained a single negative outcome, namely ‘Andrew, a fellow resident, falling into the radiator and breaking his arm’. Each vignette varied according to the extent of ‘personal control’ (high and low) displayed by Mark and the presence or absence of the mitigating factor. The mitigating factor was ‘profound communication difficulty’, which was chosen as a variable that is central to formulation of challenging behaviour in people with intellectual disabilities (Emerson, 2001). The vignettes were presented using a modified Latin square to control for order effects (Colbourn 1984). After reading each vignette, the participants completed the attributions questionnaires. Questions were included to check that intended manipulation of the levels of control (high and low) and mitigating factor (present or absent) was achieved within the four vignettes.

Measures 1. Demographic questionnaires. A questionnaire detailing participant’s date of birth, gender, length of time they had worked with people with intellectual disabilities and challenging behaviour. 2. Judgements of responsibility were measured using two items, rated on 7-point bipolar scales. Items were adapted from Corrigan et al. (2003). The first asked ‘How much is it Mark’s fault that Andrew fell over?’ and the second asked ‘How much is Mark to blame that Andrew fell over?’ 3. Participants were asked to rate their emotional response of anger and sympathy to each vignette. Two separate questions were used to measure anger; ‘How angry would Mark’s behaviour make you feel’ and ‘How irritated would you feel by Mark’s behaviour’ (Dagnan et al. 1998; Corrigan et al. 2003). Similarly, two separate questions measured sympathy, ‘How sympathetic would you feel towards Mark?’ (Corrigan et al. 2003) and ‘How sorry would you feel for Mark?’ (Weiner 1995).

Data analysis Using the logic of Baron & Kenny’s (1986) method to test the mediating effect of responsibility, we regressed the scores for emotion onto responsibility for each © 2015 John Wiley & Sons Ltd, 28, 589–593

Journal of Applied Research in Intellectual Disabilities

vignette and repeated the ANOVA using the residual scores for emotion where the variance due to responsibility had been taken out. If the interaction effect for the vignettes reduces in the analysis using residual scores, then we have demonstrated mediation. This design therefore allowed for an investigation of the mediating role of responsibility and moderating effect of mitigation central to Weiner’s (1985) model.

Results Table 1 shows the descriptive data and distribution for each dependent variable. The anger scale was not normally distributed and was excluded from further analysis. For all vignettes, the sympathy scale met the requirements for a normal distribution. Two of the responsibility scale items had Kolmogorov–Smirnov values that indicated variation from normality (Table 1), although both had Z values for skew and kurtosis that were below 2.00; on this basis, these items were included in the analysis without transformation (Tabachnick & Fidell 2007). As expected, the two items measuring responsibility were found to correlate across all four vignettes, and these items were summed to

Table 1 Descriptive data for main study variables (N = 50)

Experimental condition

Sympathy One-sample Responsibility One-sample Mean (SD) K-S test Mean (SD) K-S test

High control; 7.5 (3.6) mitigation present (Vig. 1) High control; 10.0 (2.8) mitigation absent (Vig. 3) Low control; 3.7 (2.4) mitigation present (Vig. 2) Low control; 3.5 (2.7) mitigation absent (Vig. 4)

0.71

8.1 (2.7)

0.84

1.811

6.2 (2.8)

0.91

0.95

9.2 (3.1)

1.00

1.961

10.0 (3.7)

0.99

High scores indicate high values of the named variables. 1 Distribution that differs from normal based on the Kolmogorov–Smirnov one-sample test. © 2015 John Wiley & Sons Ltd, 28, 589–593

591

produce a single responsibility scale for each vignette. Similarly, the two items measuring sympathy were also found to correlate across all four vignettes and they were also summed into a single variable for the final analysis as planned.

Manipulation check The intended manipulation of control (high and low) and mitigation (present and absent) was successfully achieved. Significantly higher mean scores in response to the manipulation check ‘How much was Andrew’s falling due to Mark’s action’ were obtained for the scenarios intended to generate high control compared to the low control scenarios. In response to the open-ended question regarding mitigation, communication difficulties were identified in 77% of vignettes where mitigation was present and in 4% of vignettes that where mitigation was absent, providing strong evidence that mitigation was well manipulated across the four vignettes.

Mediation and moderation effects The moderation of control by communication and the mediation of the effect of control on sympathy by responsibility were tested following the principles described by Baron and Kenny (1986). Before mediation can be explored, it must first be established that there is an effect between ‘moderated control’ and the outcome variable (sympathy) to be mediated and that there is an effect between ‘moderated control’ and the mediator (responsibility). To analyse this, two repeated-measures analyses were conducted. These are shown in Table 2. Both analyses have significant interaction effects, showing a moderated effect for control on sympathy and responsibility. Subsequently, sympathy was regressed onto responsibility for each vignette and the residual scores for sympathy were obtained. These scores represent the variance in sympathy that is not accounted for by responsibility. A further repeated-measures ANOVA was then conducted using the residual scores as the dependent variable. For responsibility to be seen to mediate between ‘moderated control’ and sympathy, we would expect the effects of moderated control on the residual scores to be substantially reduced, demonstrating that most of the effect of vignettes on sympathy in the previous analysis was due to the variance in sympathy accounted for by responsibility. The results of this analysis are presented in Table 3.

592 Journal of Applied Research in Intellectual Disabilities

Table 2 Tests of within-subject effects for 2 9 2 (control and mitigation) repeated-measures ANOVA with responsibility and sympathy as dependent variables Dependent variable

Independent Sum of variable d.f. squares

Responsibility Control Error Mitigation Error Interaction Error Sympathy Control Error Mitigation Error Interaction Error

1 49 1 49 1 49 1 49 1 49 1 49

F

Sig.

Partial g2

1388.6 148.0 0.001 0.75 459.6 96.6 29.7 0.001 0.38 159.6 114.0 25.5 0.001 0.34 219.2 255.4 33.8 0.001 0.40 369.6 27.38 8.3 0.006 0.14 161.6 72.00 19.6 0.001 0.29 180.0

Table 3 Tests of within-subject effects for 2 9 2 (control and mitigation) repeated-measures ANOVA with residual scores for sympathy when responsibility is partialled out as dependent variable Dependent variable

Independent variable

d.f.

Sum of squares

Residual score for responsibility on sympathy

Control Error Mitigation Error Interaction Error

1 49 1 49 1 49

281.3 3369.8 182.7 256.0 150.2 4296.4

Sig.

Partial g2

4.1

0.049

0.08

35.0

0.001

0.42

1.7

0.197

0.03

F

A comparison between Tables 2 and 3 illustrates that the removal of the variance in sympathy that is due to responsibility results in a reduction of the effect of control on sympathy (g2 reduces from 0.40 to 0.08) and a reduction in the effect size of the interaction on sympathy (g2 reduces from 0.28 to 0.03). The probability of the change in F observed for the interaction between Tables 2 and 3 is highly significant (P < 0.001).

Discussion Few studies have investigated Weiner’s later model, which describes the moderation of attributions of control by mitigating factors in the judgement of responsibility (Weiner 1995), with carers of people with intellectual disability. The findings suggest that personal

control is moderated by the apparent level of communication ability in its effect on responsibility and sympathy. This supports Weiner’s (1995) model. The protagonist in the vignettes was judged to have the same control but significantly less responsibility in vignettes where he had communication difficulties, in contrast to vignettes where he did not. As predicted, this reduction in judgements of responsibility resulted in increased sympathy. There is evidence that carers can systematically overestimate the communicative ability of adults with learning disabilities (e.g. Purcell et al. 1999) and that presenting an accurate picture of this may significantly affect attributions of challenging behaviour. Weiner’s model offers an insight that behaviour can be seen as under a person’s control but that mitigating factors can reduce responsibility judgements and increase positive emotions. It has always been a problem in this area that addressing judgements of controllability to challenging behaviour can result in observers not attributing control to any behaviour leading to an apathetic response (e.g. Woolfson 2005). Weiner’s (1995) model offers a framework within which to manage this effect in clinical practice. In summary, the findings indicate that the presence of communication difficulties moderates the effect of control on judgements of responsibility ascribed to the service user, which will in turn determine the level of sympathy in response to the behaviour. This indicates a mediated-moderation effect, which is predicted in Weiner’s later work (1995). The study has some significant limitations. The difficulty in getting a measure of anger that allows appropriate parametric analysis has been noted elsewhere (Dagnan 2012) and meant that anger could not be analysed in this study. The study uses vignettes. There is discussion of the use of vignettes in modelling the behaviour of carers towards people with learning disabilities (Dagnan 2012). For this type of study where control of stimulus is necessary to establish the fit of the model to data, the use of vignettes is appropriate; however, this research area requires more studies that demonstrate the application of cognitive models to the behaviour of staff in real settings. A clinical implication of Weiner’s (1995) model is that training and other carer interventions should focus on increasing carers’ range of causal explanations regarding challenging behaviour and encouraging them to consider relevant mitigating factors for challenging behaviour in general and for the people they work with in particular (Williams et al. 2012). This gives a possible framework for understanding the process through © 2015 John Wiley & Sons Ltd, 28, 589–593

Journal of Applied Research in Intellectual Disabilities

which assessment and formulation can affect the emotional and behavioural responses of carers. Research in this area has often suggested that carer training should support examination of attributions for challenging behaviour, and Weiner’s revised model gives a framework for understanding how it is possible to examine judgements of ‘responsibility’ for behaviour without fundamentally challenging the attribution of control, which may result in negative responses to the people exhibiting the behaviour (e.g. Woolfson 2005). The present study stresses the need for future studies to apply Weiner’s revised model (1995) and therefore include variables such as judgement of responsibility and mitigating factors. A useful further step would be to explore the extent to which different mitigating factors impact on care staffs’ judgements of control and thereby identifying which are more powerful which may help structure formulation-based training and intervention.

Correspondence Any correspondence should be directed to Dave Dagnan, Cumbria Partnership NHS Foundation Trust, Community Learning Disability Services, Unit 1, Lakeland Business Centre, Jubilee Road, Workington CA14 4HA, UK (e-mail: [email protected]).

References Baron R. M. & Kenny D. A. (1986) The moderator-mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology 51, 1173–1182. Colbourn C. J. (1984) The complexity of completing partial Latin squares. Discrete Applied Mathematics 8, 25–30.

© 2015 John Wiley & Sons Ltd, 28, 589–593

593

Corrigan P., Markowitz F. E., Watson A., Rowan D. & Kubiak M. (2003) An attribution model of public discrimination towards persons with mental illness. Journal of Health and Social Behaviour 44, 162–179. Dagnan D. (2012) Carers’ responses to challenging behaviour: a comparison of responses to named and unnamed vignettes. Journal of Applied Research in Intellectual Disabilities 25, 88–94. Dagnan D. & Cairns M. (2005) Staff judgements of responsibility for the challenging behaviour of adults with intellectual disabilities. Journal of Intellectual Disability Research 49, 95–101. Dagnan D., Trower P. & Smith R. (1998) Care staff responses to people with learning disabilities and challenging behaviour: a cognitive-emotional analysis. British Journal of Clinical Psychology 37, 59–68. Emerson E. (2001) Challenging Behaviour: Analysis and Intervention in People with Severe Intellectual Disabilities. Cambridge University Press, Cambridge, UK. Purcell M., Morris I. & McConkey R. (1999) Staff perceptions of the communicative competence of adult persons with intellectual disabilities. The British Journal of Developmental Disabilities 45, 16–25. Stanley B. & Standen P. J. (2000) Carers’ attributions for challenging behaviour. British Journal of Clinical Psychology 39, 157–168. Tabachnick B. & Fidell L. S. (2007) Using Multivariate Statistics. 5th edn. Pearson/Allyn & Bacon, Boston, MA. Weiner B. (1985) An attributional theory of achievement, motivation and emotion. Psychological Review 2, 543–571. Weiner B. (1995) Judgements of Responsibility: A Foundation for a Theory of Social Conduct. The Guildford Press, New York, NY. Williams S., Dagnan D., Rodgers J. & McDowell K. (2012) Changes in attributions as a consequence of training for challenging and complex behaviour for carers of people with learning disabilities: a systematic review. Journal of Applied Research in Intellectual Disabilities 25, 203–216. Woolfson L. (2005) Disability and the parenting paradox. Psychologist 18, 421–422.

Exploring Carers' Judgements of Responsibility and Control in Response to the Challenging Behaviour of People with Intellectual Disabilities.

This study examines Weiner's recent cognitive emotional model which makes a distinction between judgements of control and responsibility and emphasize...
105KB Sizes 0 Downloads 5 Views