Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
The Effect of Device and Mode on Measurement in Telephone and Web Surveys
Abstract
The use of mobile communication has expanded rapidly over the past decade. More than fifty percent of US adults now live in a cell phone only household (Blumberg and Luke, 2017), and close to 60% use their phones to access the internet (Pew Research Center, 2017). Survey researchers and respondents have adapted to this change, and many telephone surveys achieve at least 70% of completes from a cellular frame (Langer, 2015, Pew, 2017, Gallup, 2016), and up to 40% of completes on web surveys are from mobile devices (Saunders, 2015; McGeeney, 2015; Lugtig, Toepoel and Amin, 2016; Marlar 2018). This dramatic shift to mobile devices has raised concerns among survey researchers about the potential impact on measurement (Lavrakas et al., 2010; Kennedy and Everett, 2011; Peytchev and Hill, 2010; Buskirk, 2011; Buskirk and Andrus, 2012; Antoun et al., 2018). This dissertation builds upon previous research on mode differences to understand how mobile devices impact measurement, either within a mode or across modes. I explore three main research objectives: (1) do telephone survey responses and data quality differ across cellular and landline respondents, (2) do web survey responses and data quality differ across mobile device and computer respondents, and (3) do we still find measurement mode differences between telephone and web responses when mobile devices are included in comparisons? Overall, I find little evidence that respondents answer questions differently on mobile and traditional devices, controlling for demographics. Mean scores, probability of giving a most positive or negative response, and measures of data quality were similar between landline and cellular respondents and desktop and mobile respondents. Scale design had no meaningful impact on measurement differences between devices. Consistent with earlier work on measurement mode differences, telephone respondents give significantly more positive responses to attitudinal questions than web respondents (Dillman, 1991; Dillman et al. 1996; de Leeuw, 2005; Dillman and Christian, 2005; Brick and Lepkowski, 2008; Christian, Dillman, and Smyth, 2008). My research provides evidence that previous finding on mode differences hold with the inclusion of mobile devices. Further, there was not a scale format that effectively minimized differences between modes.
Subject Area
Social research
Recommended Citation
Marlar, Jennifer C, "The Effect of Device and Mode on Measurement in Telephone and Web Surveys" (2019). ETD collection for University of Nebraska-Lincoln. AAI13904498.
https://digitalcommons.unl.edu/dissertations/AAI13904498