Using a multi-organization database: Research methods, strengths, and limitations

Abstract

This paper describes the WFD Consulting multi-organization database and the methodology used by the five quantitative studies in this special issue to examine workplace flexibility (Hill & Civian, 2008). The database includes 72,161 responses representing 312,459 employees from 25 US organizations in five industries gathered from 1996 to 2006. These studies have an average response rate of 45%. First, an overview of the research questions is presented. Next, definitions and psychometric properties of several commonly used measures and indexes are explained. After this, analytic strategies for hierarchical linear modeling including handling of missing data, protocols for model building, and accommodating clustered data structures are described. Finally, the strengths and limitations of using this kind of dataset for academic research are discussed.

Este artículo describe los datos multi-organizacionales de WFD Consulting y la metodología usada por los cinco estudios cuantitativos de este volumen que examinan la flexibilidad en el trabajo (Hill & Civian, 2008). Esta base de datos incluye 72,161 respuestas que representan a 312,459 empleados de 25 compañías estadounidenses representativas de 5 industrias, recogidas desde 1996 hasta 2006. Estos estudios tienen un promedio de respuestas de 45%. Primero, se da un bosquejo de las preguntas. Después, se explican las definiciones, índices y las propiedades sicométricas de varias medidas de uso común. A continuación, se describen estrategias analíticas de modelos jerárquicos líneales que toleran lagunas de información, protocolos para crear modelos y para acomodar estructuras de agrupamiento de datos. Finalmente, se discuten los puntos fuertes y las limitaciones del uso de esta clase de base de datos para la investigación académica.

Keywords:

• workplace flexibility
• organizational studies
• hierarchical linear modeling

Keywords:

• flexibilidad laboral
• estudios organizaciones
• modelos lineales de jerarquías

Acknowledgements

The authors would like to thank Arlene Johnson for her vision and leadership in making this project a reality, Kathleen Christensen for guidance as a member of the advisory committee, and Joseph Grzywacz, Marcie Pitt-Catsouphes, and E. Jeffrey Hill for their collaboration and counsel in sorting out the analytical challenges of this effort. We are grateful, as well, to Ann Crouter for early project contributions and to E. Jeffrey Hill for his key role in shepherding this issue to completion. Thanks also to Donna MacKinnon, Youme Yai, and Lynn Shroder for assistance in data coding, Tina Matz-Costa for being a valued member of the project team, and Richard Congdon for technical support on software matters. This research was supported by a grant from the Alfred P. Sloan Foundation (2005-3-33).