Different survey methods are appropriate for different objectives and budgets.
Mail surveys enable researchers to get a rough sense of experiences and outcomes and are relatively inexpensive. However, welfare recipients are a highly mobile population, and many surveys may be undeliverable. The response rate of mail surveys is generally low (less than 30 percent.) Mail surveys must also be geared to the literacy level and language proficiency of the recipients.
Telephone surveys are appropriate for interviews of up to forty-five minutes and generally deliver a response rate between 60 percent and 70 percent. Although many respondents lack telephones, some studies have circumvented this problem by having field researchers provide respondents with a cell phone to call in their responses or by sending a letter offering respondents a monetary incentive for calling a toll-free number.
Mixed-mode surveys use telephone surveys followed by personal visits for nonrespondents or mail surveys with phone followup. Pursuing more than one data collection method will increase the response rate and the validity of the data.
In-person surveys use field personnel to conduct surveys in respondents' homes. Field staff can conduct longer interviews to capture detailed or sensitive information. Although this method produces the highest response rate (usually more than 70 percent), it is also the most expensive, averaging about four times the cost of telephone surveys.