Balancing MSE and Number of Contacts

Vasja Vehovar and Katja Lozar Manfreda

This paper addresses the balance between costs and errors in telephone surveys. Two alternatives are compared: a larger sample with less follow-up calls and a smaller sample with more follow-up calls. The comparison involves a detailed elaboration of mean squared errors and cost functions. Based on the model, the key variables are discussed. The empirical example refers to the percentage of companies with access to the Internet. It is shown that the above elaboration can be helpful for practical decisions.