ideoplast.org

milling ideas since 1968

Persona Reading Notes Michael Wojcik

These are my notes for the WRA 415 Persona readings (Week 4).

Maxwell & Schulman, “Personified Segmentation”

• Their method tries to combine qualitative and quantitative, be scalable, and "maintain research relevancy"
• qualitative segment: group of people who share goals; compare Agre's definition of "community". Deep description but may produce a segment that's not very relevant to the task, because no quantitative controls. Standard persona method.
• quantitative segment: group of people with similar demographic etc stats. Models how existing customers (or other target group) behave, but lacks explanatory power. Standard method for marketing segmentation.
• personified segmentation combines these approaches
• Phases:
  1. gather existing data
  2. ethnographic study - qualitative
  3. test and measure - quantitative
  4. construct and apply
  Each phase based on the results of the previous one.
• Fruit example:
  1. Research literature on fruit consumption. Figure out what you don't know.
  2. Ask people questions (what fruit do you like? what was your worst fruit experience?). Distill list of variables (eg price, flavor) from interviews.
  3. Create statistical frameworks for those variables. Use statistical interview techniques to do quantitative research on consumer fruit preferences. Analyze to group people into categories: people who focus mostly on price, people who focus mostly on flavor, etc. Discard variables that don't group well. Cluster your respondents based on similar statistics.
  4. Use clusters to determine which personas to create. Use qualitative data from interviews to "bring them to life" - add in personal details and other stuff that the statistics don't show.
• Applying personas: socialize persona use in client organization by supplying artifacts (posters, dolls, etc). (It'd be fun to have some of these things made for our personas, for our in-class exercise, but I don't know that we have time.)
• Future plans include eg pulling data from social networking to dynamically update personas.

Ford, “Creating Quality Personas”

• attitudinal / behavioral persona is based on "distinctive behaviors"
• Best personas describe how a person will go about making a key decision
• Steps for creating quality personas:
  1. identify most important research segments
  2. conduct qualitative research (Doesn't doing this for class raise human-subjects issues?)
  3. analyze data and develop personas
  (Isn't that kind of obvious?)
• Persona creation usually takes 4-8 weeks. Obviously we don't have this kind of time; part of our in-class activity might be brainstorming how to create good personas in the very limited time frame of the course.

Quesenbery, “Putting Personas to Work”

• Usability depends on the user, so personas help us study usability for different types of users.
• Some aspects of personas for website usability:
  • Role
  • Frequency of use
  • Context (eg relaxed / under pressure)
  • Choice (compulsory / independent)
  • Familiarity
  • Trustworthiness
  • Depth of knowledge
  • Relationship (eg insider, client, peer)
• Look for similarities and shared stories among users
• We have a storyteller sculpture quite similar to the one on slide 21.
• Stories show patterns of use
• "Use personas to explore scenarios" (slide 25): this could be part of our in-class activity

Calabria, “An Introduction to Personas”

• Refers to Cooper Interaction Design for background on personas. Alan Cooper is very famous in the software design field; see eg his book The Inmates are Running the Asylum. (I find some of his arguments rather unconvincing, and at times close to offensive. But he has a lot of good insights.)
• Persona research tips:
  • list groups of users
  • aim to interview around 10 users; maybe more, but stop when responses start repeating
  • interview business stakeholders
  • use market research
  • use quantitative surveys
  • anecdotal research can also be helpful
  • conduct interviews in the user's environment

Masten & Plowman, “Digital Ethnography”

• Effects of widespread ICT use ("revolution" blah blah blah) on ethography - digital ethno
• Use of ICT as anthropological "remote sensing devices" (moving up nicely on the creepiness scale here...)
• This piece isn't holding up very well on technical details - eg "imminent adoption of new Internet protocols such as IPv6". With that plus the fawning technophilia, the enthusiasm for surveillance, and the deluge of buzzwords, I have my doubts about anything coming from these guys.
• Valentine's Day study: ethnography conducted via ICT instead of by f2f interviews or immersion; participants also collected data using eg digital cameras.
• "putting the power of participant observation in participants' hands" - but this raises all the problems of self-selection, collusion, etc

Exactitudes

• So, Versluis and Yttenbroek want to show us that group identity is constructed, performed, or interpreted through clothing, or something along those lines. And?
• Though I admit some of the suits in #112 ("Sapeurs - Paris 2008") are pretty nice.