Aggregated choice data and logit models: Application to environmental benign practices of conservation tillage by farmers in the state of iowa

The many applications of soil survey and GIS based data to agricultural development and agri-environmental research necessitates the development of econometric approaches that allows researchers to combine this fine scale data with larger scale-aggregated- dependent variable choice data while minimizing the loss of information caused by the aggregation process. The study presents an intuitive, new approach to estimation of a logit model with group average choice data that are subject to a measurement error. We illustrate the approach in a Monte Carlo analysis and in an application to a dataset on farmer choice of tillage.