• Embed Doc
  • Copy Link
  • Readcast
  • Collections
  • 1
    CommentGo Back
Download
 
1
MANAGEMENT SUMMARYFrom Paper-and-Pencil to Screen-and-Keyboard: Studies on theEffectiveness of Internet-Based Marketing ResearchElisabeth C. Brüggen-Deutskens
The development of online surveys is the largest paradigm shift in theindustry since the late 1950's. This potential of online surveys is driven bytwo main developments: (1) The rapid development of the World Wide Webwhere over half a billion people worldwide now have Internet access (NUAInternet Surveys 2003) and (2) the advantages of online surveys, especiallylow costs and fast responses, which make online surveys an effectivealternative to traditional survey methods. 50% European researchers saythey have replaced face-to-face studies with online research over the lastyear (Greenfield/Ciao Online Research Barometer 2005) and 75% of researchdecision makers in the US tried online research and expect to use it again inthe future (Acorn / IIR). Some even predict that within two years time,online surveys will have an average predicted share of business of 39.6%,more than telephone (30.7%), face-to-face (23.9%) or mail surveys (5.7%)(Metzke and Allan 2005). Spendings on online surveys have increased from€3 mill. in 1996 to 500 mill. in 2002 (Hogg 2002) and the value of the onlineresearch market is expected to increase to $4 billion by 2008 (Cambiar / GMI2005).Despite the increasing popularity, scientific research on online surveys isstill scarce. This is surprising since online surveys are more than just a newsurvey mode. Online surveys present a technological and cultural changethat influences what we do as researchers and how we think about research.As the success of online surveys will stand or fall by the credibility of thedata it generates in relation to data collection costs, empirical evidence isneeded that examines critical factors and their impact on the effectiveness of online surveys. Therefore, the overall objective of my dissertation was to
 
2
investigate the impact of design factors, mode effects, and individualdifferences on the effectiveness of online surveys. The studies wereconducted in a B-2-B as well as B-2-C context using surveys, experiments,and Generalizability theory (G-theory). The types of web surveys that wereinvestigated were volunteer opt-in panels and multi mode surveys conductedwith list-based samples.This main research problem was analyzed by conducting fourconsecutive studies as outlined below.
1) Design of Online Surveys
Study one examined the impact of several important design factorswhere the evidence from the traditional survey literature could not betranslated to the online environment. The virtual setting requires, forexample, different incentive schemes since cash incentives cannot beattached to an online questionnaire. It is also unclear how a ‘long’ onlinesurvey is defined. Rosenblum, for example, states in 2001 that onlinesurveys should not contain more than 20 questions, which is generallyconsidered too short for most surveys. Furthermore, researchers could makeuse of the technical possibilities of the Internet by, for example, includingpictures. However, such a visual presentation could have an effect onresponse rates and response quality.To answer the fundamental question of how to design online surveys,we conducted an online experiment with 730 respondent to a standardizedmulti-client attitude and usage study, in which the effect of the timing of follow-ups, different incentives, length, and presentation of the questionnaireon the response rate and response quality was examined.The results showed that long questionnaires still generated asurprisingly high response, which indicates that online surveys have thepotential to be used for substantive research. Moreover, we found thatlotteries with several small prizes are very effective in increasing theresponse rate, especially in short surveys. This is surprising since the mail
 
3
survey literature finds that small cash incentives for every respondent aremore successful in increasing response rates than lotteries or donations.In addition, study one was also one of the first studies to examine theuse of design elements that make use of the interactive nature of theInternet, such as product images. Showing pictures of products instead of only giving the name increased response quality while response rates werelower, especially in the long version. Also, respondents in the visual versionpreferred an incentive that compensated for the time and effort they spenton the questionnaire rather than a donation to charity. This indicated thatrespondents in the visual version were subject to longer downloading times.
2) Comparability of Online and Mail Surveys
Once researchers know how to design online surveys, it is paramount toinvestigate the response quality of online surveys and examine whether thereare any systematic differences between online and offline self-administeredsurveys. Because of the importance of response quality, two studies in thisdissertation examined different aspects of response quality. In both studies,we conducted a large-scale real life quasi-experiment in which relevantinfluencing factors such as sample source were the same across groups sothat systematic differences between online and mail surveys could beexamined.Study two examined whether online and mail surveys produceconvergent results. Even though online surveys seem to replace telephonesurveys in the commercial world, online and mail surveys are similar instructure and are therefore often used in multi-mode service quality studies.In the context of a large B2B service quality assessment, we analyzed theaccuracy and completeness of respondents’ answers to both open and closedquestions. Overall, the results suggested that online and mail surveysproduce equivalent results. The answers of respondents did not differsignificantly, meaning that respondents were not more positive or morecritical on one of the survey modes. Also the number and length of answersprovided was similar. However, minor differences occurred in the answers to
of 00

Leave a Comment

You must be to leave a comment.
Submit
Characters: ...
03 / 14 / 2011This doucment made it onto the Rising List!
You must be to leave a comment.
Submit
Characters: ...