A free comprehensive guide for evaluating site structures
Another source of participants are the many companies that run large online research panels.
We can use a commercial panel by itself, or as an addition to other recruitment channels (to help fill out our desired numbers).
How panels work
Users sign up to these panels to earn rewards for doing online studies. When they sign up, the company collects all kinds of information about them – demographics, buying habits, hobbies, and so on. This creates a large database that can be queried for specific types of participants (e.g. women aged 40-60 who shop online at least once a month).
When we (as researchers) ask the company for participants, we can then specify which kind of people we’re looking for.
While each research panel is different, the process typically goes like this:
- We find an online panel we want to use, based on what kind of users they offer in our region, how much they charge, and whether their participant system can work with our online testing tool. This usually involves studying their website and emailing them specific questions.
- We set up an agreement with the panel company, specifying how many users we want and how much we will pay for that number of responses.
- We configure our study to receive participants from the panel. Usually this involves receiving a “participant identifier” from each incoming participant, then notifying the panel when that participant has completed the study (so the participant can earn their reward from the panel company).
Some tree-testing tools let us configure this as part of our study. Others may require us to do a bit of extra work or ask the vendor to help us set it up, or may require us to ask each participant for their identifier so we can send it back to the panel company later.
- We configure our study to return participants to the panel after the study. This informs the research company that their participant has completed our study and can be rewarded. The company should supply us a URL to use.
Quality of participants
In terms of quality of the results we get, research panels follow the same caveat as most other recruiting methods – we will always get a small percentage of participants who race through the test and give “dummy” answers, just so they can get the reward.
We haven’t found these paid panels to be any worse than other methods in this regard (and, in fact, some of them actively cull members who don’t give a decent effort, if notified about specific cases), but remember that regardless of the methods we use to recruit, we will still need to watch for garbage responses when we analyze the data later. (See Cleaning the data in Chapter 12.)
Caveats for panels
While recruitment panels can work well and can save us a lot of time and effort, there are a few questions we should ask before using them:
- Does the panel cover our region?
If we want participants from a particular province, state, or country, we need to check with the vendor to make sure they have adequate numbers in that region. International panels usually do a good job of covering North America and western Europe, but may be patchy elsewhere.
- Is the panel likely to include the types of people we want?
Most commercial recruitment panels consist of consumers. This works well if we’re targeting a portion of the general public, but it’s harder to use consumer panels for other audiences like business people, farmers, government employees, and so on. We may need to find a more suitable source of participants - see Other ways to recruit later in this chapter.
- Will we need to do additional screening of your own?
While the criteria we supply to the panel will net us a subset of their members, this may not be specific enough for our study. For example, we may have asked the panel for women aged 40-60 who shop online at least once a month, but we’re really only interested in those who have returned at least one item that they bought online.
Once we go beyond the criteria that the panel offers, we need to do additional screening – see Screening for specific participants later in this chapter.