Comments on Long Live the User (Persona): Talking with Steve Mulder

Rahel Anne Bailie — Thu, 07 Jun 2007 03:32:03 GMT

In response to Boke’s comment re a “set of smart questions,” a methodology has been developed that does delve deeper and draw out more meaningful and pertinent information. It’s now known as Cognitive Edge, and was pioneered by Dave Snowden (formerly of IBM’s Cynefin Centre, or is that of IBM’s former Cynefin Centre). I was so impressed with the methodology that I got my accreditation this spring.

At any rate, here’s an example of the qualitative difference of data that can be gathered. A survey asks on a scale of 1 to X, say 1 to 7 for purposes of this exercie, about a matter, with the best rating being the higher number. How people interpret the question will affect how they average the rating. For example, “how do you rate the administrative function” could rate 4 if it’s average, or a person could rate it 4 if they kind of like 3 things but detest 1 thing. And the more surveys you gather, the less meaningful the data becomes. However, if you gather anecdotes (the quality of the prompting questions are critical), and then attach have the respondents rank the anecdotes (in effect, using metadata) on a bell curve where answers in the mid-range are acceptable, and the “trouble signal detection” lies in the outer ranges, then it becomes more obvious where the clusters are. First, you gather information about things that may never come up on a survey; second, because the respondents do the ranking, the attribution ranking is accurate; and third, the good/bad dichotomy is replaced by a hypo-state/normal-state/hyper-state paradigm that is more tolerant of what is “normal.”

Michael Beavers — Fri, 23 Mar 2007 16:32:11 GMT

Boke, I highly recommend reading Mulder’s book if you’re interested in persona development or learning about how user research is used to inform IA. It is vastly different from police interrogations and psychology that attempts to uncover subjects’ deception. It isn’t a threatening environment—or at least shouldn’t ever be—so the cumpulsion to lie isn’t really an issue. While cognitive psychology does play a role, these methods draw more heavily from cultural anthroplogy and related field studies.

If user or persona research is designed well, research participants do not intentionally mislead the researchers. That said, however, what people say they do versus what they actually do can be very different. This is why a balance between quantitative and qualitative is critically important. ‘Psychological tests’ don’t really gain the researcher anything. The purpose of user-centered design is to inform IAs and interaction designers so they can build systems and interfaces that aid the user in their pursuits. It is a group of methods that actively and iteratively involve users in the design process.

If you’re new to information architecture, you’ll probably also hear people frequently refer to the “Polar Bear Book”, which is Information Architecture for the World Wide Web by Rosenfeld and Morville. This is a critical, if long, primer for the IA professional—read it!

Boke L Nkoy — Thu, 08 Mar 2007 05:45:14 GMT

I am new to IA. Listening to experts talking about personas makes me think of one thing. Why can’t some forms of ‘psychological tests’ be designed for application during users’ research? To build good personas, we need to collect quantitative and qualitative information. As concerning qualitative analysis (or interviews), what you want is the user to tell his or her frustration about using the website, reasons for that frustration, etc. In most of the cases, it will be embarrassing for the person to let you know his or her own ignorance about use of computers and websites.

In this case, I don’t see how methods such as surveys will help to get at the bottom of this problem. Probably, observing users using the computer may be a useful way, but then, you can only see what he or she is doing; but the mental process of this information search is hard to guess. Of course, observation allows picking up a few clues about usability problems, but still may not yield much in terms of what goes on in the mind of the users.
Someone has to come up with a “psychological test applied to usability’, identify a set of smart questions that could tease out subconscious answers, discriminate a lie from the truth and retrieve what really goes through the user’s mind. Such a test may yield necessary elements to build more realistic personas.

In my view, personas are close to “identikit pictures”. Police officers always build them so as to put a face on their offender. It is amazing that with few details, the officers are able to make a drawing close to the real face. Identikit pictures should held us to understand how to make better personas.

Boke Landu Nkoy
I am studying Information Architecture at the University of Canberra. I am finding all those subjects fascinating. I came from a different background; a bit of statistics and economics. So, I enjoy all those writings; unfortunately, I start to realize that IA, personas, controlled vocabularies, diagrams, content analysis, taxonomy, labeling, etc seem to be new fields of studies where virtually we are still trying to invent the wheel.

Paula Thornton — Fri, 02 Mar 2007 17:23:12 GMT

I interact with (or see the back side) of a lot of companies. Sadly, those who even have dedicated resources aligned to the web site are far fewer than those who do (what they ‘say’ and what they ‘do’ are far different behind the scenes).

Michael Beavers — Thu, 15 Feb 2007 01:38:36 GMT

Liz, the link to his book site and Chapter 3 were a great bonus to a good article. What I like about Mulder is that he is from the camp of “reasonable persona” development that seems to recognize that companies have research cultures with varying appetites for qualitative versus quantitative methods.

Company marketing and customer insights departments have for so long relied almost exclusively on quantitative for critical business decisions, while qualitative research has been regarded dismissively as something that “marketing and advertising does with focus groups”. Within companies this has caused an enculturation of a limited world view of qualitative research.

Now that that the lines between mass-media marketing/advertising and product development are blurring, there is a wonderful opportunity for more holistic views of customers, the messaging to which they will be receptive, and what they will actually interact with. This is a very exciting time to be in research, design, strategy, and brand planning.