Evidence-Based Practice: Attitude and Knowledge of Social Workers across Geographic Regions

Abstract

The objective of the author in this article was to examine possible differences in knowledge and attitudes toward evidence-based practice (EBP) among social workers across geographic regions. A random national sample of 180 NASW members was obtained from mail and Internet groups. MANOVA analysis was performed to determine possible differences in knowledge and attitudes toward EBP among these social workers. Findings suggest that knowledge and attitude toward EBP did not differ among these practitioners. Despite increasing efficacy and widespread knowledge of EBPs, there is little or no empirical evidence to support any differences in attitudes and knowledge of EBP among social workers across geographic regions.

Keywords::

• Evidence-based practice
• knowledge of EBP
• attitude toward EBP
• EBP knowledge across geographic regions

Acknowledgments

The author would like to thank Donna Harrington, PhD (Professor), Llewellyn Cornelius, PhD (Professor), and class members for their helpful comments and contributions to the success of this research. Many thanks also go to the University of Maryland, Baltimore (UMB) School of Social Work for funding and supporting this project. In addition, the contributions of all experts who responded to the pilot survey for this project are gratefully acknowledged.

Notes

^1. The author considered controlling for age and conducted a Multivariate Analysis of Covariance (MANCOVA) but the assumption of linear relationship between the covariate (age) and the DVs (attitude and knowledge) was not adequately met. Thus, there was no linear relationship between age and knowledge of EBP (F = 2.887, p < .91), and age and attitudes (F = 2.209, p = .139). “Nonlinear analysis of covariance is possible though it is rarely done” (Huitema, 1980, as cited in Stevens, 2002, p. 347). Therefore, MANOVA was performed.

^2. The formula K(K-1)/2 was used to determine the number of Hotellings T2s, and the number of t-test(s) was estimated by multiplying the number of DVs (2) by significant T² (Stevens, 2002).