Definitions of Community: 

 An Illustration of 

 Aggregation Bias 



by 

 A. E. Luloff and P. H. Greenwood* 



INTRODUCTION 



Aggregation belongs to the class of empirical problems which 

 may be assumed away. While assumptions tend to be painless, they 

 are normally not costless. The presence of aggregation problems 

 may bias one's results and distort one's hypothesis tests regardless 

 of whether or not they are assumed away. Potential problems with 

 aggregation arise for many reasons. One important cause is a 

 research interest that defies consensus definition. For example, 

 there are researchers interested in the behavior of community and 

 while these researchers are prepared to offer a definition of a com- 

 munity, they are hard pressed to find an operational analog. Sim- 

 ilarly, marine economists may consider models of a fishery when a 

 fishery is no less abstract a concept than a community. If a county is 

 regarded as a collection of communities the problem of defining a 

 community is avoided provided that aggregation problems are 

 assumed away. Those familiar with the quasi-community litera- 

 ture will not be surprised at the number of data collection units 

 which have been used as either community surrogates or aggre- 

 gates. In the absence of any aggregation problems, we should be 

 indifferent as to the unit over which data is collected. Simply assum- 

 ing a problem away is not the most satisfying method for dealing 

 with a problem that poses real hazard of distoring one's perception. 



The Problem of Aggregation 



Applied research on the nature of community typically involves 

 the empirical determination of the relations among a set of varia- 

 bles, and the comparison of these results with theoretically derived 

 hypotheses. It should be clear that, if the comparisons are to be 

 meaningful, the empirical and theoretical results must be compara- 



*Assistant Professor of Community Development and Assistant Professor of 

 Resource Economics, respectively, Institute of Natural and Environmental Resources, 

 University of New Hampshire, New Hampshire Agricultural Experiment Station, 

 Durham, NH. 



