Development of a Scale for Nurses to Assess Their Own Shift Leadership Behavior

Abstract

We have located few systematic studies that use existing scales for shift leader nurses who are responsible for nursing in a team while receiving instructions from physicians, coordinating and directing team members during specific shifts. The present study uses a development method different from existing scales to reduce the number of items. This study aims to develop a scale for nurses to assess their own shift leadership behavior, and establish the reliability and validity of the scale. A 37-item draft scale was created based on the results of an anonymous open-ended questionnaire administered to shift leader nurses working in hospitals in metropolitan areas in Japan and a literature review. Using the draft scale, an anonymous self-rating questionnaire survey was conducted with 631 shift leader nurses working in hospitals in metropolitan areas in Japan. As there was a statistically significant difference between male and female nurses in the mean total scores of the draft scale, we decided to analyze only the 336 female nurses. Based on principal factor analysis and promax rotation, a scale for nurses to assess their own shift leadership behavior that contains 25 items in 6 factors was developed. The six factors are “task coordination behavior”, “human relationship building behavior”, “effecting collaboration through information transmission”, “confirmation of the performance of the nursing practice”, “information sharing behavior in teams”, and “development of nursing skills”. The goodness of fit of the hypothetical model obtained by an exploratory factor analysis was confirmed by a confirmatory factor analysis, and we obtained following results : GFI＝0.880, AGFI＝0.850, RMSEA＝0.053. The correlation coefficient with the Self-evaluation Scale on Role Performance of Shift Leader Nurse was 0.786, a value which has shown criterion-related validity. Cronbach alpha coefficients were 0.94 for the whole scale, and 0.78-0.88 for subscales. The intraclass correlation coefficient by the test-retest method was 0.88. We decided that these results support the reliability. The findings show that the reliability and validity of the scale comprised of 25 items in 6 factors were largely established.