Measuring implicit associations with behaviours to improve resident mood: development of implicit association tasks for nursing home care providers

Abstract

Objective

To develop and evaluate instruments for measuring implicit associations of nursing home care providers with behaviours aimed at improving resident mood.

Method

Study 1 (N = 41) followed an iterative approach to develop two implicit association tasks measuring implicit attitude (positive versus negative valence) and motivation (wanting versus not wanting) regarding mood-improving behaviours, followed by an evaluation of the content validity for target stimuli representing these behaviours. In Study 2 (N = 230), the tasks were assessed for stimulus classification ease (accuracy and speed) and internal consistency. A subsample (n = 111) completed additional questionnaires to evaluate convergent validity (with self-reported attitudes towards depression, altruism, and mood-improving behaviours), and discriminant validity (against social desirability), and repeated the tasks after 2 weeks to assess test-retest reliability.

Results

Content validity indexes for target stimuli were satisfactory. Error rates were acceptable for attribute stimuli, but exceeded the 10 % limit for target stimuli. Response times for all stimuli exceeded the 800-millisecond threshold. Both tasks demonstrated good internal consistency but poor test-retest reliability. Regarding convergent validity, both tasks significantly correlated with altruism, the implicit attitude task associated with self-reported mood-improving behaviours, and the implicit motivation task correlated with the behavioural scale of attitudes towards depression. Discriminant validity was supported as neither task was significantly associated with social desirability.

Conclusions

The implicit association tasks show potential for measuring implicit associations with mood-improving behaviours of care providers, offering an innovative pathway for exploring processes influencing caregiving behaviours. However, limitations in psychometric properties were identified, aligning with challenges observed in similar measures.