Article
UCD 2/5 Analysis and Understanding

User Research

Observation

Through observation we collect behavioural data. Other sources of behavioural data are for instance: Logfiles, eye tracking, physiological analysis and usability tests (for more informations on usability test visit the last episode of my UCD series).

Types

• free observation; is explorative, there are no directives
• partial-standardised observation; the focus depends on the questions
• standardised observation; is often used for evaluation

Methods

• open observation; user know that they are being observed
• covert observation; the users don't know that they are research subjects
• participating; the analysts are part of the observation
• not participating; the analysts are not part of the observation

Questioning

Through questioning we gather statements and attitudes of users. A general problem is the low correlation between the users satisfaction and objektiv performance. The truthfulness of self-disclosures is also an issue.

Interviews

Interviews can be conducted with and without an existing system. With a system we want to find problems and without we want to find ideas. The order of questions is oriented on a natural conversation flow. An advantage over questionnaires is the possibility to intervene if misunderstandings or uncertainties occur. The number of interviews depends on the type and size of a project but not less than 4 is recommended.

Possible disadvantages

• extensive procedure
• falsified answers through suggestive questions
• social pressure on the interviewee

All the more expertise interviewees have, all the more problems they get with explaining what they are doing and why. That's because of tacit knowledge, humans don't think about the steps of their tasks or always know them explicitly. An other aspect is that users mostly don’t say what thy really think and there are many reasons for that. Therefor we have to recognise ambiguities and watch the body language.

Ineffective interview styles

• I ask, you answer
• intimidating
• indifferent
• small talk
• focused on the schedule
• expressing new ideas or solutions

Best practice examples

• introduce yourself and explain the purpose
• use an open formulation and no suggestive or multiple choice questions
• be open to adjust your schedule
• document everything and not only the answers
• give it a trial run

Questionnaires

Types

• ad-hoc questionnaire; is system and task oriented and the structure is of importance
• closed questionnaire; predefined number of possible answers
• partial open questionnaire; leaves the user the ability to add own text
• open questionnaire; no predefined answers, the user writes free text

Advantages

• more anonymous
• time flexible
• standardised or standardizability
• scalability
• automatic evaluation

Possible disadvantages

• preparation depends on prior knowledge
• low response or return rate

Contextual Inquiry

Is a mixture of observation and questioning conducted while a user is on site working. The goal is to learn what the users are doing, what is important for them and how they are structuring their work. Finding out the work details, that users don’t talk about because they are obvious for them or they do them unintentionally plays a big part.
The execution is based on the master and newbie metaphor where the analyst becomes a trainee that learns from the user.

Phases

• traditional interview; introduction and so on
• Transition; watching and asking questions
• Contextual Inquiry; observation and asking questions
• Wrap-Up; feedback and interpretation

Task and Problem Analysis

The scope of task and problem analysis include

• ongoing processes
• logical and temporal dependences
• frequency and urgency of tasks
• critical fault situations

There are numerous methods for analysing and partly they build upon each other, the choice of methods depends on the tasks and problems.

Hierarchical Task Analysis

Tasks are separated step by step into subtasks and elementary activities. The hierarchy is illustrated with nodes for every subtask and build from left to right on chronological order.

Personas

Gives a fictional user a face, vita, family, hobbies, office, colleagues and so on. It is based on real informations and represents a typical user from the target group.

Szenarios

Describes use or problem scenarios from the users point of view. It is suited for discussions with users to verify the correctness and relevance of aspects. A scenario should contain at least a person, relevant surroundings, goals and actions.

User Stories

Describes requirements from the users point of view and can be drafted on demand. They are written on story-cards which contain a identifier, a descriptive title and a functional description. Template: As “role” I will “goal”, because “justification”.

Use Cases

Illustrates a series of interactions by the user with the system and the response of the system. It consist of a diagram and a textual description.

Context Analysis

The target is to understand the context on all levels. The scope typically includes the technical and organisational context.
Example questions that need to be answered can be:

• Where is the system used and on which platform is it based?
• How is the system used and who is allowed to do what?