Screening

Screening Types

When the sample frame is subsectioned or matched against an external administrative data source, this is referred to as passive screening because a respondent is not directly involved in the process. Passive screening uses existing data to determine who, from a sampling frame of individuals, establishments, or other, is likely eligible for a survey. For instance, a survey of pediatrie specialty hospitals in the western United States may begin with a list of all hospitals across the United States. Based on the original list itself, or another that has been merged with the original, the list can be narrowed down to those hospitals located in western states. The list could be further screened to include only those with a pre-designated classification of being a pediatrie hospital.

When eligibility information is obtained directly from a respondent or proxy, this is referred to as active screening. Active screening involves direct contact with potentially eligible respondents and is typically undertaken when the eligibility criteria are not available from the sample frame. In this scenario, potentially eligible respondents are contacted in person, by phone, by Web, or by mail to determine their eligibility through a short screening interview (or “screener”). A variety of eligibility criteria may be evaluated by actively screening respondents. Considerations may include age, race, education, income, or geographic location, among others, depending on the purpose of the survey. Active screening is also often done to identify rare or difficult-to-locate populations. For example, a household survey of Hispanic single mothers may include an active screening component to determine the age, ethnicity, and relationships among all household members to make a determination as to whether anyone in that household is eligible as “Hispanic single mother.” Households would be contacted and asked questions related to these demographic characteristics, and only those meeting the eligibility criteria would be retained for possible participation in the survey.

Active screening can be completed using several different modes and at a time different from data collection for the main survey. Often, but not always, the screening takes place in the same mode as the main survey interview. For instance, in a general population household survey, an interviewer may visit the household in person to administer the screener to the person answering the door. The screener may be designed to identify household members with a certain characteristic or within a certain age range. From the screener results, the main interview respondent or respondents can be selected as specified by the sampling criteria from the pool of eligible household members, and the interview can then be conducted in person with the selected respondent or respondents. Screening can also take place over the phone, Web, or by mail, which are often more cost-effective techniques compared to in-person screening, though each has associated sources of potential error.

...