Page 85 - Reduction of coercive measures
P. 85
Causal attributions
Support staff causal attributions on challenging behavior were measured with the Dutch version of the Challenging Behavior Attribution Scale (CHABA, Hastings, 1997; Zijlmans, Embregts, Bosman, & Willems, 2012). The questionnaire contains 33 items on a 5 point Likert scale (-2 very unlikely and 2 very likely). The questionnaire distinguishes five causal attributions: learned behavior (six items, three items for learned positive and three for learned negative behavior), medical/biological factors (six items), emotional factors (seven items), aspects of the physical environment (eight items) and self-stimulation (six items). Examples of items are ‘because he/she cannot cope with stress’ (emotional cause), ‘because he/she lives in a noisy place’ (physical environment cause), and ‘because he/ she wants attention of people’ (learned behavior). Subscale scores were determined by summing all scores of the five causal models and dividing them into the amount of items belonging to a model. A low score, below zero, indicated that the respondent considered this cause unlikely as an explanation for challenging behavior and a high score the reverse. Hastings (1997) showed acceptable to good internal consistency values for all of the CHABA subscales, Cronbach’s alpha varied from .65 to .87. The present study aimed at testing hypotheses based on the theory of Weiner (1985, 1986). Therefore, the items of the CHABA were transformed into the subscales stability and controllability using computations of the studies of Bailey et al. (2006), Lambrechts et al. (2009), and Zijlmans et al. (2012). Cronbach’s alpha for the scales controllability and stability was respectively .80 and .85. Scores of support staff within a care unit were aggregated to a mean score. A moderate Pearson correlation of .74 was found. Because the two scales measure two mutually exclusive sub-aspects of attribution, the two scales were retained separately.
Difficult Behavior Self-Efficacy Scale
Self-efficacy of support staff was assessed by the Dutch version of the Difficult Behavior Self Efficacy Scale (DBSES) (Hastings & Brown, 2002; Willems, Embregts, Hendriks, & Bosman, 2016). The five items were presented digitally and rated on a 7-point Likert scale. Items
Associate factors of coercive measures
83
4