Online Program

Do Different Correspondence Methods Improve Response Rates for Out of Scope Cases?

Keywords: Likely out of scope, Autodial, Postcard, Multiple modes, Response rates

For many surveys, some of the sampled units are not members of the population of interest. Because the survey appears irrelevant to them, these units often do not respond indicating they are out of scope for the survey. This discussion focuses on the likely out of scope records for the US Agricultural Census for which the agricultural activity is unknown. This group generally has low response rates as most do not qualify as agricultural operations, but they are still required to return the form indicating their status. Non-response from this group increases costs for follow up contacts and may increase non-response bias. The National Agriculture Statistics Service (NASS) experimented on the 2010 Census of Agriculture Content Test and on the National Agriculture Classification Survey (NACS) with different methods to encourage likely out of scope respondents to return their report forms. NASS compared its standard messages to a combination of cover letters, postcards and reminder messages stressing to respondents that they were required to return the form even if they were not involved in agriculture and if they did not respond they would be personally visited or telephoned.