R. Chris Fraley How to Conduct Behavioral Research over the Internet: A Beginner's Guide to HTML and CGI/Perl (2004) Guilford Pres.
MAY/JUNE 2004
Volume 10 - No. 2
SRNT Newsletter
May/June 2004, Volume 10, Number 2
Book Review
In recent years, the Internet has shown increasing promise for the recruitment of research participants from an enormous and diverse population. As a growing proportion of the general population continues to gain access to the Internet, concerns about accessing underrepresented populations have been receding. The limitations of the use of this resource for large-scale data collection appear now to often reside with the researchers themselves, who may lack the technical expertise and time required to develop online experiments or who may lack funding for programmers who could assist with this. Consequently, the growth in Internet-based research has not kept pace with the increased accessibility to and use of the Internet by the general population.
How to Conduct Behavioral Research over the Internet is described as "one-stop shopping for any behavioral scientist thinking of creating a website to collect data". The book begins with a general overview of the basic materials and knowledge that will be required in order to collect data on the Internet. It then leads the reader gently through using a web server, designing simple HTML web pages (i.e., HyperText Markup Language, for those of you, who, like me, did not know this!). It is not until Chapter 4 that the technical details associated with collecting data on the Internet are revealed, with plenty of examples and screenshots to guide the more nervous programmer. CGI / Perl is also introduced, which enables more interactive data collection than HTML.
Subsequent chapters demonstrate how to provide automated feedback to participants and enable desirable characteristics of experiments such as randomization (of both stimulus presentation and group assignment), tracking over multiple sessions, measurement of response times, and so forth. It is these later chapters which really make the book stand out as a genuine resource for behavioral researchers. Simply placing questionnaire measures on the internet is relatively undemanding but also often inadequate to answer many important research questions. The ability to conduct experimental research studies on the Internet has the potential to vastly increase the scope and impact of such an enterprise. For example, there are inherent difficulties in using the Internet to measure response times if this is influenced by the speed of the userÕs Internet connection. Solutions to such dilemmas are clearly addressed.
However, while the book has considerable merit, it would have been enhanced by more attention to the scientific and ethical challenges presented by Internet data collection. For example, researchers have reason to be concerned about issues such as who is taking part in the experiment, and whether their data is accurate or even honest. Although ethical issues are discussed in the final chapter, this coverage is rather cursory. For example, different research review committees may have different views on the ways in which participants are able to withdraw during an experiment. Is it sufficient to simply assume that participants can log off if they so choose, or should an explicit opt out option (e.g. a visible button on the website running the experiment) be made available?
Overall, however, these limitations are minor issues, and should not detract from an excellent, informative and, above all, timely introduction to data collection on the Internet Š an introduction which should be accessible to all but the most technophobic of behavioral researchers. I tried the worked examples and, to my mild astonishment, they worked. The online population of the world is now available to me and my research program Š I intend to make use of this rich resource immediately.
Marcus Munafò, Ph.D., is a Cancer Research Fellow in the Department of Clinical Pharmacology at the University of Oxford in Oxford, England