Research shows that using a prenotice letter is effective in increasing survey response. By informing respondents that their address was selected to participate in a survey and providing information regarding the survey, prenotices help address legitimacy and confidentiality concerns. However, prenotices are one of many pieces of survey materials mailed to respondents, therefore, their value needs to be considered carefully from the perspectives of respondent contact burden and costs.

Much of the research on prenotices has focused on traditional data collection methods — paper, telephone and personal visits. However, surveys are moving away from traditional methods and toward the web. With this in mind, now the question is whether prenotices are still as effective as they have been in the past. If so, what are other ways to increase survey response and still minimize respondent burden and survey costs?

Last year the U.S. Census Bureau conducted the 2015 Mail Contact Strategies Modification Test to see if we could increase self-response to the American Community Survey by modifying our mail materials and contact strategies. The presence of a prenotice was one of the items tested. We tested this alone and in combination with other modifications, such as changing the format of a later reminder from a postcard to a letter.

We used several treatments to test various materials and strategies to see which worked best with the American Community Survey’s push-to-web data collection method. This method is designed to initiate response to the American Community Survey through the internet and then follow-up with nonresponders via paper, telephone and personal visits. Our goal was to increase self-response via the internet and mail, which are the least expensive modes. We compared self-response return rates to see which strategy best met our goal.

The test concluded that simply eliminating the prenotice did not have a significant impact on total self-response return rates. However, we found that we could eliminate the prenotice and increase internet response rates by sending the initial mailing earlier and replacing our first postcard reminder with a letter reminder. Both the initial mailing and the letter reminder provided the user ID information for respondents to respond via the internet. It appears that earlier access to the Internet instrument allows more time and increases internet response. The internet return rate for the treatment with the combined modifications (no prenotice and letter) was 3.5 percentage points higher than the internet return rate for the control treatment with the prenotice and postcard reminder. In addition to the significant increase in self-response return rates, the study estimated substantial cost savings for conducting the American Community Survey.

The success of the test prompted the Census Bureau to implement the new mail contact strategy into the American Community Survey production beginning with the August 2015 production panel.

For more information, see the report 2015 Mail Contact Strategies Modification Test.

We are also presenting our findings at the American Association for Public Opinion Research’s 71st Annual Conference in Austin, Texas, during the first poster session on May 12, 2016.