[Scenario-Based Qualitative Research _UX1]

[Scenario-based qualitative research _UX1]

Take customer needs and reflect them in the intent of what you build


Understand the user experience (how) / understanding -> with a strategy / strategy -> reflect it in design (how) / design
Set goals > analyze the context > field research > modeling   >     UX strategy   > service design  >  prototyping  > test & build



Goals: market selection, business context understanding, defining the UX challenge, reviewing the business model
       -> STP strategy for market selection,
       -> Understand business goals (first grasp the causes of upstream concerns, then set the UX direction: trends, value networks, etc.)
       -> Define expected effects -> organize the business model

Context: statistical analysis (raise issues/phenomenon - but you can't know the cause; find that through research),
          UX evaluation (heuristic, AHD, etc. to analyze usefulness), trend analysis, WOM analysis
      -> Analyze the current state from current data -> analyze the trajectory from past to present -> type analysis (process and categorize data)
      -> Spot patterns, derive implications, analyze issues and severity (cultural / loose error checking, serious error checking)
      -> Expert evaluation (will a low score actually agree with the user's perspective? / 5- or 7-point scale)
      -> Trend analysis: societal and cultural shifts—where people are getting excited, where they've lost interest—and talk to individuals
      -> WOM analysis -> derive key findings

Field research: verify the phenomenon/issue, research plan, choose research techniques
       -> Define the proposal and the scope precisely, then
       -> User (type, traits, needs and motives), brand strategy/design, online channel strategy
       -> "Hypothetical persona" -> research plan—target selection (classification table): 1-2 days, define the method (card sorting, observation, video): 1-2 days,
           content design (field research guide): 1-2 days, recruiting and schedule: 4-7 days -> "real person"
       -> Picking research subjects: classify users inside the target market -> gradually segment each (occupation—office, self-employed... / preferred category—fashion, beauty... / marriage—kids...)
       -> Define the method
           * Interview techniques (individual/group)
              - Card sorting (reading the mental model; you can hear the language they use directly from them, and label it)
                            (open sorting: hand them information and let them categorize and read it back...)
                            (Delphi card sorting: one person goes first and others revise, making access and participation easier)
                            : various language, styles, attitudes come out, but you can find the common threads
              - Cognitive map (usually in the latter half of an interview): in what context do you use it, what value do you want, distinctive questions,
                                                    what's the motive for using the product? Link these motives to draw out behavior patterns.
                                                    Write down the hurdles blocking that flow.
                            : Participants' hurdles are actually more about drawing pictures/circles.
                              Design around that; pick experienced people over novices.
              - Be your customer: talk to people (clients) whose job is interacting with end users
              - Word concept association: language represents that person's emotional territory (David Aaker's language map)
              - Unfocus group interview: gather people from contrasting segments.
                     : Focus group interview—can be risky. People are bound to engage in social validation.
                       Especially with an authoritative expert, a friend, or a similar person around, they end up ignoring small differences and agreeing easily.
             - Extreme user interview: talk to people who seem unlikely to ever use it
                                                  Insights come from people outside the typical user range.
                                                : It gives you another lens to reinterpret the general findings.
             - Long range forecast: in group interviews, people often offload opinions to others; show a video matched to the service line/industry,
                                              and you can get good feedback from the group. (Great for preparing future-facing options.)
             - Design the box: like cognitive map, a late-interview technique
                                        Besides transcripts and video, use an actual box and let them write the key points themselves: features / region name
                                       -> The form triggers action and makes it memorable.
             - Free listing: what there is matters, but what's important also matters a lot.
                                (Cultural anthropologists use it to grasp the cultural traits of a society or tribe)
                                (Sometimes people stay quiet in the interview, then raise issues after the sorting exercise)
         * Observation methods (scenario-based qualitative research): environment of use, performance capacity / user traits, scenarios, key issues
             - Photo diary: have users capture their own daily life in photos
             - Role playing: have the user role-play as the provider ("what would you do?") / participatory design
             - Shadow tracking: centered on a specific person, specialized B2C for sales reps, insurance agents, etc.
             - Town watching: centered on a specific place (subway: 20s use maps and apps; 50s just maps)
                                      (Observation points: flow, before/after payment, after pickup, people count: alone/two/three-four, common patterns)
             - Error analysis: people tend to blame themselves for errors' causes.
             - Collage: show images representing various types and let them pick;
                            but insight comes from understanding why they picked the image, not just the image itself.
             - Personal inventory: belongings, apps
             - Activity analysis

             - Scenario: portal -> search "Jeju Shilla" -> instead of hitting shortcut, some pick images or blog posts first
                             Keep them talking about their situation.
          * 5 why

Modeling: organize the results, affinity, persona, user story
       -> Organizing the results
            - Non-verbal expression beyond words also matters a lot (ethnography for marketers): intonation, volume, attitude toward me
            - Notes, recordings, cam, artifacts
            - Organize daily, and separate fact from opinion
            - Manage important issues separately
            - Organize by issue
            - Organize by context (by place, how time affects experience, situation, social/cultural factors)
       -> Persona: pick a specific persona as the primary user and build service traits and design around them.
                     At this point, the language from card sorting can be a strong source of inspiration.

Strategy: value definition, value review, strategy formulation

Service design: service ideation, concept model, user scenario, mental modeling

Prototyping: storyboard, paper/hi-fi prototyping

Test: execution plan, execution, result analysis (errors and time are relatively easier to diagnose than emotion, etc.)

Implementation: service feature definition, design





* Validity assessment and reliability judgment on research results (a quantitative attitude?)
  vs qualitative research is about finding the purpose and cause / meeting the right people matters most.

* Even with qualitative analysis, the people reading it tend to understand it quantitatively.
  : hide the formal parts and unpack the story around the issues (content).

* Before sharing or commissioning survey results,
  first clarify the 'why' behind the issue, then run the survey and present.

* Something new
  - Grounded in context, lifestyle, and usage behavior.
  - In the end, the user isn't a new person.
  - Do qualitative research first, before quantitative research.