3. Defining Objectives
Surveys are used to collect information that will answer scientific questions. The first consideration in designing a survey is to define the objectives of the survey in the context of these questions.
The objectives should reflect the intent of the scientific questions from the perspective of the survey methodology to be used to collect data for addressing these questions.
Survey objectives should specify:
- The study population (also called a target population) and any subgroups (also called domains) that are of special interest in addressing the scientific questions
- The organization of the population (e.g., housing units are associated with a group of people, students are nested in schools), which often has implications for gathering information on units in the population
- Likely methods for contacting members of the population or gathering data on selected units
- The main variables that will be recorded on the units of the population to address the scientific questions
- Available resources such as limitations on staff funding and time for conducting the survey
Using this information, a design is developed to meet the scientific objectives of the survey, often via an iterative process that leads to modifications of the design and possibly the objectives. Think about a population you would like to study and describe a variable you would like to measure. What are the units of measurements? What would be a high value? What would be a low value? What would be a typical value? What kind question(s) would you ask to obtain data for this variable? Make a note of these, to consider later in this chapter.