|
Researcher FAQ |
|
|
|
|
|
|
|
|
|
|
|
|
|
Here are some answers to questions researchers frequently have about the StudyResponse project. Prospective participants should visit the Information for Participants page.
Why should I collect data using StudyResponse?StudyResponse was created in response to difficulties that we and our colleagues had while trying to recruit participants for various types of studies that are not sensitive to so called "non-sampling errors." In particular, StudyResponse is likely to be useful for studies of phenomena that exhibit robust correlational relationships. StudyResponse has also been used for qualitative data collections, where data would not be analyzed with statistical techniques. Generally speaking, StudyResponse is inappropriate for demography and other applications where coverage errors could adversely bias your results. How does StudyResponse work? You craft a sample of participants to receive your participation request. Next, we send recruitment email messages to the individuals in your sample. Participants link to your survey by clicking on an URL in the message. Except in prescreening studies, data management and maintenance of the survey page are accomplished by you. Click here for more detailed information on the StudyResponse research process. We are not a "vendor" of email addresses: Our work with you is part of our research on the characteristics of studies (such as length of survey) and study outcomes (such as response rate). How will I develop my survey and who will host it? The StudyResponse project provides no survey development or hosting services, except for prescreening studies. The best surveys (i.e., most attractive and easiest to use) tend to be custom developed by a professional web developer. There are also many commercial services for developing web surveys and these range in cost from free up to thousands of dollars. Some StudyResponse researchers have used SurveyMonkey, which provides a very limited basic service for free and more expensive services for a monthly charge, and SurveySuite, which offers a free trial followed by a yearly account charge. Here are vendors who offer software packages or services that can assist with web development: Assess:
(http://www.assess.biz/) How much does it cost to use StudyResponse? StudyResponse is an academic, non-profit research project dedicated to improving the feasibility of online research. The web-based facilities and databases for the system are generously hosted by the School of Information Studies at Syracuse University; administration is accomplished by a team of researchers and research assistants. To support these processes we use a percentage of the overall cost of participant incentives provided by researchers. Please see our guidelines for additional information. What kind of panelists does StudyResponse have? We have a varied sample consisting of many tens of thousands of people, from many occupations, racial/ethnic groups, and ages. Many of these are U.S. residents although we have an increasing number of international panelists. Have a look at our sample characteristics page to see more detail. Will my study be publishable using this kind of sample? The short answer is: "Yes, assuming that you have other sources of data and/or theoretical justifications for using a sample that contains coverage bias." The long answer is that we have a bibliography of studies that have been published using online samples. To help you in your effort to get your paper published, we have developed technical papers that you can cite as evidence for the quality of the data obtained through StudyResponse. Remember, though, that the likelihood of publishing of your study depends upon many factors other than the source of the respondents. In particular, your choice of theory, methods, measures, and analyses will have a powerful influence on the overall quality of your study. What kind of response rate can I expect if I use StudyResponse? Response rates for studies done using the panel have varied from as high as 80% to as low as 2%. In general, as the panel has grown larger, response rates have declined. At present it is not atypical for a researcher to obtain a response rate between 10-30% using a 7-day reminder. Higher response rates are possible using by conducting a prescreening study first or by offering direct payment (as opposed to random-draw) incentives. Not unexpectedly, response rates depend on the completion length of your research instrument, the topic of your study, and incentives offered for participation. Preliminary results from our research suggest that attractive surveys under 10 minutes in completion time and incentives configured as a sizeable number of modest prizes tend to garner the highest response rates. Sadly, we cannot guarantee any specific response rate for your study, and we cannot provide a refund if your study fails to reach a response rate you consider appropriate. What
other studies have been done with StudyResponse, and how
have they turned out? As
of September 2005, researchers from four countries (U.S., U.K., Canada, and
Australia) have conducted 218 studies using
StudyResponse. Most of our researchers are "repeat customers" who conduct
2-4 studies per year. Our technical reports
page refers to some published work that has used the StudyResponse
project. What kinds of studies does StudyResponse support? StudyResponse supports many different kinds of studies. They all have the common characteristics of being based on self-report, using web-based data collection, being anonymous and confidential, and being approved by a registered Institutional Review Board (or equivalent body at international universities). Within these parameters a wide range of methodological possibilities exist for qualitative and quantitative data collection, multiple waves of data collection, experimental manipulations, and so forth.
|
