Dimensionality and Reliability of the Japanese Version Emotional Response Questionnaire

Tuesday, 14 July 2009: 8:30 AM

Chiemi Kochinda, RN, PhD
School of Nursing, University of Kansas Medical Center, Kansas City, KS

Learning Objective 1: Understand dimensionality of the Japanese version Emotional Response Questionnaire.

Learning Objective 2: Assess reliability of the Japanese version Emotional Response Questionnaire.

Purpose: To examine dimensionality and reliability of the Japanese version Emotional Response Questionnaire (J-ERQ).

Methods: Applying forward and back-translation techniques, translation equivalence of the J- ERQ was established prior to data collection.  Data was collected from 158 nursing students in Japan and analyzed using SPSS 15.0J. Truncated principal component factor analysis with criteria of factor loading ³ .40 was conducted using Varimax rotation. Scree plot analysis and Eigenvalues > 1.0 were used as the criteria for factor identification.  After the final number of factors was determined, each factor in the J-ERQ was assessed for its reliability using Cronbach’s coefficient alpha.

Results: The analysis results yielded an 12-item scale with two dimensions: the Distress and the Empathy subscales, which coincided with the original theoretically based dimensions. Together, these factors accounted for 45.3% of the variance. Eight distress items constructed the first factor and accounted for 30.2% of the item variance with the Eigenvalues 4.52. Factor loadings in the Distress subscale ranged from .58 to .82. Four empathy items constructed the second factor and accounted for 15.1% of the item variance with the Eigenvalues 2.27. Factor loadings in the Empathy subscale ranged from .48 to .84. Item analysis in this factor revealed that the item “Sympathy for the person” appeared to have inadequate homogeneity with the other items. Its inter-item correlations were .10-.32, the corrected item-total correlation was .24, and alpha-if-item-deleted increased by .07. Consequently, rewording of this item or elimination from the subscale was suggested. The theta reliability estimate for the whole scale was .85 and Cronbach’s alpha for the Distress and the Empathy subscales were .88 and .70, respectively.

Conclusion: The J-ERQ resulted in two dimensions, the Distress and the Empathy subscales and their reliabilities are established. However, there is a need for concept clarification between empathy and sympathy as nursing students tend to perceive them differently.