@dana > > Just get a mix of participants who do and are motivated to do what you're > interested in observing. > agree.....
If one or two participants in your mix have the issue, you want to fix that > because you don't want *anyone* to have it. > Totally agree in the ideal world all issues should be fixed. With Webnographer some of our clients are so overwhelmed by a long list of issue, what they are wanting is prioritisation of issues. Many times there is a very long list of usability issues that have been waiting for years in some cases. By being able to say X% of users experienced this issue starts helping the solution to be prioritised, and fixed. Often there is an argument between one camp and another camp if a usability issue is real or not. Prioritisation it seams is very important to Agile operations. James http://blog.feralabs.com 2009/3/12 Dana Chisnell <[email protected]> > > Ah, I meant with regard to age. If the sample is 8, say, in 99.9% of cases, > age won't matter. Just get a mix of participants who do and are motivated to > do what you're interested in observing. > If you're testing 30 or 50 or 100 participants, you might want to pay > attention to make sure you have participants from all the age ranges you > care about, but you shouldn't be selecting or screening on age as long as > they do and are motivated to do the same kinds of tasks with your designs. > > Most demographics don't matter in usability testing, most of the time. (Of > course, there are exceptions.) Why? Because the purpose of usability testing > is not to generalize preferences to a larger audience but instead to > identify problems with a design that cause frustration and confusion. If one > or two participants in your mix have the issue, you want to fix that because > you don't want *anyone* to have it. > > Dana > > > On Mar 12, 2009, at 10:21 AM, James Page wrote: > > @dana > I am bit confused here by your question "What difference does it make how > many you're testing?" > Surely factors such as "margin of error", and "statistical power" > are important, or are they not? > > The point of testing is to find out if your wrong, or right. How do you > know if your wrong or right based on a small sample. > > @jenrandolph > On remote usability testing we get more behavioural differences by machine > configuration, then by age. What I mean by machine configuration is > manufacture and screen size. Mac users are different, why - I don't know. > And we get allot of behavioural differences by culture - (place of birth vs > residence). Also environment seams to have quite a large impact. People in > the lab, and at home spend more time to trying to complete a task before > giving up, then people at work. This is of course impacts success/failure > rates. > > We are doing more research here. > > James > http://blog.feralabs.com > > > 2009/3/12 Dana Chisnell <[email protected]> > >> >> What difference does it make how many you're testing? By breaking the >> sample into groups, you're just creating extra work. Are you going to >> compare the data by age group? Why would you do that? The only reason I can >> think of is if you're creating different sites. You're not. >> Dana >> >> >> >> On Mar 12, 2009, at 9:21 AM, James Page wrote: >> >> Out of interest how many participants are you testing with? Could >> you break the numbers down? >> James >> http://blog.feralabs.com >> >> 2009/3/12 Dana Chisnell <[email protected]> >> >>> >>> Thanks for the prompt, Jared. There's no reason to limit the age range >>> *at all.* As long as the behaviors are the same -- that is, the task goals >>> of the users -- across age ranges, then it doesn't matter a bit how old the >>> participants are. >>> >>> As members of UPA, people over 65 would very likely have the same tasks >>> and goals in mind as someone younger: Maintain membership information, renew >>> memberships, find out what's going on in the association, get in the >>> consulting directory, find out who is on the board, find out where the >>> conference is, etc. >>> >>> Limiting the age range wouldn't benefit the research. In fact, limiting >>> may be a detriment. >>> >>> Dana >>> >>> :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: :: >>> :: :: :: >>> Dana Chisnell >>> desk: 415.392.0776 >>> mobile: 415.519.1148 >>> >>> dana AT usabilityworks DOT net >>> >>> www.usabilityworks.net >>> http://usabilitytestinghowto.blogspot.com/ >>> >>> >>> On Mar 12, 2009, at 12:04 AM, Jared Spool wrote: >>> >>> So? Why limit the age range? How does that benefit the research? >>>> >>>> On Mar 11, 2009, at 10:20 PM, Todd Zaki Warfel wrote: >>>> >>>> Perhaps because the core audience isn't older than 65? Not to say that >>>>> there aren't any, but I'd imagine, based on the meetings and conferences >>>>> that I've been to, that the number of people over 65 are statistically >>>>> quite >>>>> small. >>>>> >>>>> On Mar 11, 2009, at 11:54 AM, Dana Chisnell wrote: >>>>> >>>>> May I ask why the age range limits to 65? >>>>>> >>>>> >>>> >>> ________________________________________________________________ >>> Welcome to the Interaction Design Association (IxDA)! >>> To post to this list ....... [email protected] >>> Unsubscribe ................ http://www.ixda.org/unsubscribe >>> List Guidelines ............ http://www.ixda.org/guidelines >>> List Help .................. http://www.ixda.org/help >>> >> >> >> > > ________________________________________________________________ Welcome to the Interaction Design Association (IxDA)! To post to this list ....... [email protected] Unsubscribe ................ http://www.ixda.org/unsubscribe List Guidelines ............ http://www.ixda.org/guidelines List Help .................. http://www.ixda.org/help
