How to Do Mail Surveys in the Digital Age: A Practical Guide

Abstract

Mail surveys present opportunities for reaching out to specific populations in ways not achievable by Internet or telephone, but the logistics of mail surveys can be complex and unclear. This article details the post-design and pre-analysis phase of a mail survey, from printing through distribution, with particular emphasis on time requirements and cost. Preparing a mailing of 10,000 pieces manually required 240 labor hours, excluding background research, or about 1:25 per unit (pre-printed envelope with pre-printed postage, stuffed with pre-printed and hand-stamped inner envelope, booklet, and folded cover letter). Manual preparation resulted in savings translating to an estimated equivalent hourly

Every Door Direct Mail in US survey research: An anonymous census approach to mail survey sampling

“…The total costs are reported based on external material and service costs, including printing, office supplies, paper goods, postage, and services like translations. Costs associated with time and provision of "in kind" services like storage space are not included, which is consistent with other literature but masks a large hidden cost (see Grubert, 2017a for details on time requirements). Costs are divided evenly across returned completed responses in a group and reported separately for mailings done by EDDM only, addressed mail only, and addressed questionnaires accompanied by EDDM postcards.…”

“…This means that if researchers cannot drop off mailpieces at the appropriate DDU, the mailpieces must be shipped to that DDU at the conventional package shipping costs. Especially given minimum mailing sizes of either 200 units or the total number of addresses served by a given DDU, bulk shipping costs can be large (Grubert, 2017a). Additional time is also required for processing: in this study, the time between bulk shipping and actual mailpiece distribution was generally 3-5 mail days.…”

“…Preparation, however, is simplified because address labels and prepaid postage indicia can be printed directly on mailpieces (Figure 1). This benefit can be large: for a 10,000 household, three-contact study, not needing to apply stamps and labels to each mailpiece saves an estimated 3 weeks of full-time work (Grubert, 2017a).…”

“…Mail surveys are a well-established mode of reaching potential respondents in a given location (Groves et al, 2011). Mail surveys face limitations, however, including cost and time requirements (Grubert, 2017a). Reliable methods for surveying the general population at relatively low cost are widely sought (Battaglia et al, 2016), particularly given the declining response rates over time (Connelly et al, 2003;Curtin et al, 2005).…”

The Need for a Preference‐Based Multicriteria Prioritization Framework in Life Cycle Sustainability Assessment

“…Data collection for the proof‐of‐concept study described in this work cost about US$0.05 per pair‐wise comparison in payments to participants alone. Results from a broader mail‐based survey (with a goal of reaching populations difficult to recruit online) assessing 30 impact categories suggest costs of between US$20 and US$40 per response for much lower resolution data, relying on Likert‐type items rather than pair‐wise comparisons (Grubert ). Anticipated costs for a broader study addressing a full complement of about 30 impact categories (necessitating 435 pair‐wise comparisons without fatigue, a major cognitive challenge), using a sample size of about 100,000 people globally to capture rare perspectives and provide sufficient resolution to assess cultural trends, would likely be millions of dollars, not including costs associated with recruiting representative samples, convening data quality verification meetings, or analyzing the data.…”