ABSTRACT
This chapter discusses the survey as a means of collecting useful information for evaluating public programs. The essence of evaluation is comparison, and surveys are one (but not the only) way to collect information useful for comparing programs or for comparing individual performance in a special program to individual performance in agencies or administrative units without the special program.1 Thus, surveys are a particularly useful source of data for cross-section experimental or quasi-experimental designs or for cross-section nonexperiments with statistical controls. Surveys are less likely to be useful as a source of data for time-series designs. For example, in evaluating education programs (or other social service programs), administrative data can be used to compare performance in similar programs (e.g., attendance rates or test scores). However, often it is useful to supplement that data with information from surveys of participants (e.g., students or their parents) or program case workers (e.g., teachers), especially when that information is not available else where.2 (It is not a good use of resources to use surveys to collect information that is already avail able elsewhere.) The purpose of this chapter is to explain how to design a survey and use the results as one source of information to assess the effectiveness of a public program.
