In self-administered surveys, respondents must navigate through the questionnaire and interpret what each question is asking rather than assisted by an interviewer in a face-to-face or telephone interview. For an individual question, sound visual design helps respondents proceed through the response process. Visual designs are well tested by various researchers for their influence on reducing the measurement error and item non-response. In addition, Christian et al. (2007) and Christian (2007) found that adding a verbal instruction to the question’s visual design significantly improved item response rates. Research on effect of clarifying instructions on visual design is self-administered surveys is, however, limited. In order to get further understanding of the use of clarifying information in self-administered survey questions, we designed an experiment which examined the effect of clarifying instructions (with/without) on responses to two numerical open-ended questions. Data for the study was collected using two questions from the 2015 client survey of Florida Cooperative Extension Service (FCES). A web/mail mixed-mode design yielded 1,618 responses (51.4% RR2). The two questions used in the experiment were asked information about number of times using FCES this year, and years using Extension services, and clarifyinginstructions for two questions were “include 1 for the time that this survey asked about” and “write ‘1’ if only this year” respectively. We found consistent evidence the added instruction clarifying how to respond reduced the percentage of missing and incorrectly formatted responses by 4.6 percentage points for the number of times clients used Extension and 7.5 percentage points for the number of years. Response mode did not affect these results.