• Create BookmarkCreate Bookmark
  • Create Note or TagCreate Note or Tag
  • PrintPrint
Share this Page URL
Help

Part: III Appendixes > Choosing Between-Groups versus Within-Groups Experimenta...

Choosing Between-Groups versus Within-Groups Experimental Design

One important consideration in experimental design is whether each participant participates in more than one experimental condition. Say, for example, that you're testing two versions of a Web site to compare them. In this scenario, there are two experimental conditions. In a between-groups experiment, you break your pool of test participants into two groups and each group uses only one of the Web sites. In contrast, in a within-groups experiment you have only one group of test participants, and each participant uses both sites.

These two types of experimental design have trade-offs. For example, a within-groups experiment may not require as many test participants before producing statistically significant results. If you're after bottom-line data, the within-groups approach can save you considerable time and money. On the other hand, a within-groups experiment can raise issues of validity if learning effects are involved. For example, if you test the same tasks on two versions of the same Web site, your participants might be quicker completing a task the second time because they learned how to do it on the first site. You can alleviate some of these problems by randomizing or counterbalancing the order of sites tested and other experimental conditions.


PREVIEW

                                                                          

Not a subscriber?

Start A Free Trial


  
  • Creative Edge
  • Create BookmarkCreate Bookmark
  • Create Note or TagCreate Note or Tag
  • PrintPrint