The Total Survey Error Approach: A Guide to the New Science of Survey Research by Weisberg Herbert F

Author:Weisberg, Herbert F.
Language: eng
Format: epub
Publisher: UniversityofChicagoPress
Published: 2010-11-22T16:00:00+00:00

Easing the Technological Burden

A special concern for Web surveys is the extent to which unit nonresponse is due to technical problems. The respondent’s computer hardware or software may not be able to handle the questionnaire. Browser problems and slow connect times may discourage survey participation. These situations may improve over the years as browsers are improved and standardized and as fast connections become more common, but researchers have to remember that many respondents do not have Internet capabilities that permit the same multimedia access that the researchers themselves have.

The respondent may also lack the technical skills that the survey assumes. Dillman and Bowker (2001) point to several such factors that lead to nonresponse in Internet surveys. First, some people do not know how to provide answers, as in how to handle a drop-down menu. People may also not understand how to change answers, since some formats require only choosing a different answer whereas others require clicking a second time on the original answer. It is awkward when not all the answer options show up on the screen at the same time so that viewing the options requires scrolling up or down the page. Also, sometimes it is necessary to take several actions to answer a question (such clicking on an answer and then having to click elsewhere to go on to the next question).

An early problem with Internet surveys was that respondents could not tell how long the survey would be. If people got the impression that the survey was endless, many would break off early. Therefore it has become common to show respondents how far along they are in the survey through progress indicators, such as a bar in a corner in the screen which darkens as the person gets further through the questionnaire.5 Allowing people to skip questions rather than requiring each question to be answered before the respondent can go on to the next one can also decrease breakoffs.

An unusual danger of Internet surveys is that respondents seem to be especially sensitive to the survey administration. A 2001 study by the Council for Marketing and Opinion Research (CMOR) found that Internet respondents tended to be more critical of the mechanics of the survey than were telephone respondents (Business Market Research Association 2002). Experience in being members of Internet panels can further exacerbate this tendency. One result is that people may self-select out of Internet surveys that do not satisfy their view of how such surveys should be presented.

Some of these factors involve item nonresponse as well as unit nonresponse, but respondents who get discouraged by factors such as these are probably less likely to complete the questionnaire and to participate in future Internet surveys. Skilled programmers can prevent some of these potential problems, and standards are developing for others, but differences in respondents’ familiarity levels with computers will inevitably persist. Therefore it remains essential to seek to ease the technological burden for respondents in Internet surveys.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.