environment in which things get done
Figure 2.1.
This User Research Types matrix will help you decide which research to use.
Generative research has been around for a long time, though its application to software strategy is more obscure. Generative research explores the mind space of someone doing something. It is research that is conducted before the ideation phase.[2] It focuses on a higher level than evaluative research, asking the end purpose for every tool used rather than the details of how well a specific tool is applied. Open-ended research methods, such as ethnography, non-directed interviews, and diaries, allow researchers to create a framework based on data from participants. This framework then can be used to guide information architecture, interaction design, and contextual placement of information and products.
For example, you would use generative research to find out how people buy books, which usually doesn’t differ based on age, gender, interests, or preferences. You would use preference research to find out which book a person would buy. However, the act of buying is not that different: Look for the book, make sure it’s the right choice, and then purchase it. Mental models generated with generative data and aligned with proposed information and functionality can deliver an unambiguous picture of how well a solution supports the user through gap analysis.
Building Products Based on Preference Research is Like Building a Kitchen from a Stack of Magazine Clippings
Imagine you are an architect talking to a couple who wants a kitchen remodel. They bring you a stack of magazine clippings with photos of kitchens they like. They talk about how they want to cook gourmet meals for friends with fresh produce from the farmer’s market. Your next step is to start drafting blueprints based on all this input, plus the information you already know, such as the efficiency of a work triangle in a kitchen. But imagine if you were not allowed to draw blueprints, and instead you were required to hand over the stack of magazine clippings to the contractor. Assume the contractor has never built a kitchen before, so he has no experience with kitchen functionality and work triangles. Without your skills at interpreting client input into a remodeling plan, the project would stall.
Asking engineers to build a product based on a stack of preferences is just like asking a contractor to build a kitchen based on magazine clippings.
Stepping back to a more general, rather than user/consumer, definition of research, Wikipedia defines three main forms of general research methods:
Exploratory research, which structures and identifies new problems
Constructive research, which develops solutions to a problem
Empirical research, which tests the feasibility of a solution using empirical evidence[3]
In a way, mental models embody both exploratory and constructive research, allowing you to derive solutions to problems from the data set as well as structuring where new problems for the next year might lie.
How Mental Models Hook into Other UX Techniques
If you are who I think you are, you’re familiar with standard user-centered design techniques, such as writing scenarios based on specifically designated personas. You’ve probably seen or used affinity diagrams that show group relationships among things. You may have commissioned or participated in a field study of your users. You possibly have directed “Voice of the Customer” surveys as a part of a Six Sigma[4] program at your company. But you’re looking for something to pull these techniques together, to make them reach further.
Mental model research occupies a place in the constellation of techniques after user data collection and before product and interaction design concepts (see Figure 2.2). Its use as a planning roadmap is long-lived. You can refer to the same mental model for several projects over time.
Figure 2.2.
Constellation of some user-centered design steps. (No wonder it seems so hard to figure out where to start!)
Mental models are also useful for things other than design. Sales and customer service can use the data to understand clientele better. MBAs and information designers can re-format the data into workflow and process diagrams. Project managers can use it to prioritize among a set of development options. I encourage you to reach out to these people and introduce them to any mental models that you create.
Within the realm of designing solutions, mental models provide a nexus for all the other tools in your toolbox. You can draw benefits from using mental models to support your personas and scenarios. Mental models along with web analytics and use cases influence your interaction design concepts. Prototypes coming out of these concepts undergo usability testing to touch base with the user.
There are a few additional techniques that could flow directly into or out of a mental model. I’ll sketch these additional techniques here, and then dive into the main techniques later.
Input: Diaries
Diaries are a popular way to gather data in the user’s voice. You could ask participants who are, for example, members of Weight Watchers to write down their daily successes and frustrations with the diet and exercise program they are trying to follow. You can comb through this data for behaviors and create a mental model from this analysis. Diaries do have a tendency to flit from subject to subject, however, without deep examination of a topic. This tendency can leave you with a spotty mental model. But if you have this data, go ahead and mine it for your mental model.
Input: Field Visits
Field visits conducted by a professional researcher produce a much deeper understanding, and all topics within a scope are likely to be covered. In this respect, field visit data is better for task analysis than diary content. To create a mental model of the participant’s perspective, however, you will need to convert third-person notes into first-person behaviors. This translation is not insurmountable, and you will have a solid mental model as a result.
Output: Personas
According to Alan Cooper and Robert Reimann in About Face 2.0: The Essentials of Interaction Design, personas are user
