Date of Graduation


Document Type


Degree Name

Doctor of Education (Ed.D.)


School of Education


Learning and Instruction


Learning & Instruction EdD

First Advisor

Robert Burns

Second Advisor

Mathew Mitchell

Third Advisor

Steven Morris


The purpose of this study was to examine the construct validity of social loafing using convergent and discriminant validity principles. Three instruments that purport to measure social loafing were factor analyzed: A ten-item instrument by George (1992), a 13-item instrument by Mulvey and Klein (1998), and a 22-item instrument by Jassawalla, Sashittal, and Malshe (2009) for a total of 45 items that were compiled into a single instrument with which data were collected, correlated, and factor analyzed.

One hundred and sixty graduate and undergraduates enrolled in management courses at a small private Northern California university were surveyed. Thirteen classes were surveyed and data was collected over three semesters.

Data collected were factor analyzed using Principle Axis Factoring and rotated using Promax with Kappa = 4 for each instrument. Correlations, Keyser-Meyer-Olkin, and Bartlett’s test of sphericity were inspected for reasonable factorability, sampling adequacy, and appropriateness of running a factor analysis. Eigenvalues > 1 and Scree plots supported the number of factors extracted with primary factor loadings of .4 or higher. Pattern, structure, and factor correlation matrices were inspected for content, loadings, and correlations among the derived factors. Derived factors were compared to each author’s theoretical framework. Additionally, the eight derived factors were factor analyzed using the same procedures. The result was three final derived factors.

Findings showed correlations among the author’s scales indicated that the three instruments do not measure the same thing. George’s and Jassawalla et al.’s instruments share 55% of the variance. Mulvey and Klein’s instrument shares little in common with Jassawalla et al. and virtually nothing with George. Further, George, Mulvey and Klein, and Jassawalla et al. had hypothesized10 scales whereas my factoring had eight factors. Findings showed that the 8-factor solution supported George, partially supported Mulvey and Klein, and did not support Jassawalla et al. The final 3-factor solution does help to define the social loafing construct. The findings suggest using the instruments with caution. Further research to ensure accurate conceptualizations of the social loafing construct should be continued.