How to choose the right user experience research method

Note: This article is published in the Web site of usability guru Jakob Nielsen , which describes how to choose the right user experience research method. Good research methods to use the right place at the right time to play its due effect, do not have a hammer in hand to see what is like nails. Now use the research group according to our company's situation recompile revise under, for everybody reference.

User Experience Research can now answer quite a wide range of questions. By listing a variety of available research methods in a 3-D coordinate system and a typical product development phase, you can learn when to use which method.

There is a fairly wide range of research methods in user research, which is a good thing (and probably a bad thing). Methods that have been widely validated, such as usability laboratory studies, to those developed in recent times, such as consensual (desirability) studies (to measure aesthetic needs)

You can't apply all of the methods to all projects, but most design teams benefit from the insights that many research methods combine. The key question is when and what should be used. In order to better understand when and what research methods to use, it is helpful to classify various research methods according to the following three dimensions:

• Attitude and Behavior
• Qualitative and quantitative
• The background of the website or product use

The following diagram describes the positions of several common methods in the coordinate system

Each dimension is a way of distinguishing between different studies that answer different questions and are suitable for different kinds of purposes.

Attitude-Behavior Dimension

The difference can be summed up in what people say and what people do (often quite different). The purpose of attitude research is often to understand, measure, or learn about people's specific perceptions, which is why attitude research is often used in the marketing sector.

Although most usability research should rely more on behavioral research , the information obtained using the self reporting method is still useful. For example, card sorting research can give you a deeper understanding of the user's mental model in some kind of information space, which can help you decide what is the best information architecture for your site. Survey methods measure attitudes, or collect self reporting data, to help track or identify important issues on your site. For a variety of reasons, the focus group's role in usability research has begun to decline .

At the other end of the axis, the research methods that focus on behavior are often used to try to understand what "people do" and to minimize the interference of research methods themselves on research results. AB Test is only the design of the revised website, but try to keep other factors unchanged, so as to easily observe the impact of web design on user behavior, Eye movement research is used to understand the user and web Interface design visual interaction.

Between the two extremes are two of our most common research methods: Usability Lab Research and Field field research . They combine self reporting and behavioral data, and can be biased towards either end of the axis. But the general recommendation tends to be in the section of behavioral research.

Qualitative-Quantitative dimensions

The basic difference between the two is that in qualitative research, data is often collected directly, whereas in quantitative studies data is collected indirectly through a tool, such as a questionnaire, or Web server log. In field research and usability lab studies, for example, researchers directly observe how users use technology (or are not) to meet their needs. This gives them the ability to ask questions, to detect behavior, or to tweak research programs to better meet their goals. The analysis of the data is often not very accurate.

In contrast, the insights in quantitative research typically derive from accurate mathematical analysis, since data collection methods, such as survey tools or Web server logs, capture massive amounts of digitally encoded data.

Because of their differences in nature , qualitative research methods are better suited to answer questions about why or How to solve a problem . On the contrary, quantitative studies can do a better job of answering the number of questions and how many . The chart below depicts how the previous two dimensions affect the kind of research methods that can answer questions.

Product Use background dimension

The final distinction to make is whether to use the product or how to use the product in the study. Can be described as follows.

• Use products naturally or almost naturally
• scripting use of products (used in a prearranged fashion)
• do not use the product in the study
• mixing of the above items

When choosing to use products naturally in research, the aim is to minimize the impact of the study itself on results so as to be aware of real behavior or attitudes as much as possible. Many ethnographic field studies are devoted to this, but there are always some observational biases. Interception of Access surveys (Intercept Survey, translators are not quite sure that this is the way to translate) and data mining/analysis techniques are examples of such use of products in quantitative studies.

scripted Research for product use (scripted study) is designed to focus on very detailed situations, such as when redesigning a process. The degree of scripting can be quite diverse based on different research goals. For example, a benchmark study (benchmarking study) is often fairly rigidly scripted, so it can produce fairly reliable usability standards .

Studies that do not require the use of products are used to examine issues that are broader than use and usability, such as branding research and cross-cultural behavior.

Hybrid research approaches use products in an innovative form to achieve goals, for example, participatory design allows users to interact with design elements, rearrange those design elements, and discuss why they make that choice.

Most of the research methods in the chart can be moved on one or more axes, and in some studies two directions are average, often to meet multiple goals simultaneously. For example, field studies can focus on what people say (ethnographic interview ethnographic interviews) can also focus on what people do (expand the observation extended observation) (desirability) Research and card classification have both qualitative and quantitative versions, and eye movement research can be scripted or scriptable.

Phase of product development (Time dimension)

Another important difference to consider when choosing a research approach is the product development phase and its associated goals.

• Planning Phase: in the first phase of product development, in general, you need to consider new ideas and opportunities for the future. The research approach at this stage is quite diverse.
• optimization phase: Finally, you will reach a "dry or not" intersection, you transition to a period, during this period, you will be in the direction of your choice to constantly improve their design. At this stage, the main research approach is structured and helps you reduce the risk of execution.
• Evaluation phase: at some point in time, the website or product is used by enough users, you can start to evaluate how you are doing.

The following table summarizes these goals and lists the typical research pathways and approaches, and their links.

 	product development phase
Planning Phase	tuning Phase	evaluation phase
Goal:	Inspire, explore and choose new Directions and opportunities	Learn and optimize your design to reduce risk and improve availability	Measure product performance and compare it to itself and its competitors
Research approach:	Qualitative and quantitative	primarily qualitative (structured)	Mainly quantitative studies (aggregated)
Typical methods:	Ethnographic field studies, focus groups, journal studies, questionnaires, data mining and analysis	Card classification, field research, participatory design, paper prototyping and user experience research, desirability (desirability) research, customer mail	Available benchmarks, online evaluations, surveys, AB tests

Art or Science?

Although many user experience research methods come from the practice of scientific research, their goals are not pure science and need to be adjusted to meet the needs of stakeholder (stakeholder). This is why the descriptions of these methods are very brief (just general guidelines) rather than strict classifications.

Finally, the success of your work depends on how much you can achieve by upgrading your website or your product's user experience. The meaning of this classification is to help you make the best choices at the right time.