Page 431 -

422 CHAPTER 14 Online and ubiquitous HCI research

14.2.4 ONLINE RESEARCH DESIGN CHALLENGES
14.2.4.1 Appropriate topics for online research
Although it may seem somewhat obvious to note that online research will involve
working with participants who are online, this helps point us toward the insight that
online HCI research may be most appropriate for studies about the tools that people
use online and the uses that they make of those tools. Participants in online studies
will probably be working with web browsers, chat tools, and related online software
as they read instructions, provide informed consent, perform tasks, and otherwise
complete your experimental protocol. Research that works within this realm may be
most successful.
Specifically, studies involving web applications or online tools are particularly
well suited for online research. If you are running the web site on your own serv-
ers, web logs (Chapter 12) can provide useful feedback regarding timing, tasks, and
errors. Conversely, studies of other application software, mobile devices, or novel
interaction devices may be harder to do online: data collection is likely to be more
difficult, incompatibilities between software versions may pop up, etc.
That is not to say that online studies of web site designs are easy. Good de-
sign practice certainly calls for cross-platform testing, but there is no guarantee that
you will not run into versioning and compatibility problems, even with seemingly
straightforward web pages.
14.2.4.2 Recruiting
By opening your research up to the Internet, you provide yourself with access to a
much larger pool of participants. Recruiting can be easier, as emails to appropriate
lists and postings on various web sites can go a long way toward identifying poten-
tial subjects. As online research generally involves the use of a web site or other
online software, participants do not need to be local. Self-driven web site or study
tools allow participants to complete tasks at their leisure, eliminating the need for
scheduling.
Just as the use of undergraduates as study participants introduces a bias that may
not be appropriate for some studies, online recruitment limits your subject pool
to a particular segment of the larger population: Internet users who are interested
enough to participate. This may mean that you might not attract relatively inexpe-
rienced individuals or participants who limit their time online to relatively focused
activities. Whether or not this poses a problem depends on the specifics of the study
in question.
In some cases, online research can give you access to pools of participants that
otherwise would have been unavailable. This is particularly true for people with dis-
abilities, who may find traveling to a researcher lab to be logistically unfeasible
(Petrie et al., 2006), and domain experts, who may be hard to find in sufficient num-
bers in some locales (Brush et al., 2004). See Chapter 16 for more details on HCI
research involving people with disabilities. Collaborative research involving dis-
tant partners can also be substantially aided by online tools for communicating and
gathering data.