Mail versus Telephone Surveys: Potential Biases in Expenditure and Willingness-to-Pay Data

Authors

  • Doug Whittaker
  • Jerry J. Vaske
  • Maureen P. Donnelly
  • Darla S. DeRuiter

Keywords:

Mail surveys, phone surveys, recreation expenditures, willingness to pay, user fees

Abstract

Economic data obtained from recreationists are an increasingly important component of natural resource management and decision making. A major methodological issue is deciding whether to collect this information through mail or telephone survey modes. This paper compares mail and phone surveys for collecting information about Colorado state park visitation, income, annual capital expenditures on outdoor equipment, and willingness-to-pay (WTP) marginal increases in daily/ annual park entrance fees.Previous research comparing mail and telephone surveys suggests the two modes may provide contrasting results because of (l) sample representation issues and (2) context issues. Sample issues can be further distinguished into sample frame or coverage bias(whether certain people in a population are not included in a sample frame because they are unreachable) and sample non-response bias (differences between respondents and non-respondents who were contacted with each survey mode, but refused to answer the survey or items in the survey). Either bias may be partially addressed by weighting responses by sociodemographic variables if those are known for a population.Considerable research has compared mail and phone surveys regarding context issues: the effect of survey type on how people interpret and answer questions. While a variety of effects and biases have been defined, three appear relevant to recreation economic information: social desirability and acquiescence, cognitive demand effects, and response order effects. Based on this research, social desirability effects appear to have the greatest potential to affect willingness-to-pay variables, while cognitive demand effects are more likely to affect recall variables such as expenditures or frequency of visitation.Results show that although the mail survey respondents had slightly higher incomes, the two survey modes provided statistically similar results for park visitation. When income differences were controlled, expenditure differences were also statistically similar. In contrast, mail respondents reported a lower willingness to pay for park entrance fees than phone respondents.These findings suggest greater social desirability influences with the phone survey for the willingness-to-pay variables, as well as greater levels of strategic responses with the mail survey. In contrast, minor differences between modes for expenditure and visitation variables indicate that cognitive demand effects are similar for these survey types. Results imply that choice of survey mode could have effects on willingness-to-pay information (phone surveys will show greater WTP).

Published

1998-07-03

Issue

Section

Regular Papers