Usability has the following five attributes:
The "usability testing" is to invite a group of real typical users to operate the product for a typical scenario, and encourage them to try to complete the task when they think out loud, the usability staff on the sidelines to observe, listen to, record, and find out that the product is used in the process of usability problems. Usability staff observes, listens and records to find out the usability problems in the process of product use. It is adapted to the various stages of product design and development, midterm improvement and late maintenance, and is an important manifestation of the idea of user-centered design.
We as a product Manager, Designer, Coder, as a product of the parents, the child has a few moles are very clear, the child's personality with me no problem, that this function as a matter of course should be so designed. So the product designer is an expert user, and the user knowledge of ordinary users, we can not fully represent the ideas of typical users. Listen to the real voice from the user, in favor of improving the design value.
So usability testing is mainly to measure product usability from five dimensions: ease of learning, efficiency, memorability, error rate, and satisfaction, and at the same time, to locate the problems and causes of the product, and then optimize the product.
Another point is that design needs to balance the relationship between the product and the business, so usability testing is also closely related to the company's business purpose, so that testing can be an effective tool for solving problems and seeking opportunities.
NNG's Jacob Nielsen lists the following three scenarios:
1. During an iteration, especially between two iterations. We need to know if our current design solves the previous problem, or if we need to continue to improve our design solution.
2. Data doesn't lie - for usability analysis of competing products, the metrics obtained from usability testing can be very useful.
3. We need to have a clear goal in mind before each new release. A usability test can tell us if a new release is ready for release when we are not very sure about the current design solution.
The focus of the problem for usability testing needs to be defined and needs to be as precise, clear, measurable, and observable as possible, e.g., about function points, interfaces, flows, etc. Stakeholders can be asked to understand what they want to know, focusing on research objectives that will achieve product ROI.
First of all we need to have the **** knowledge that "usability testing tasks are for the actual goals of the user, not what I think the user wants to do". Secondly, here are a few core methods of task design:
a) Make a list of tasks based on the purpose of the test, the tasks should not be too many, and must be closely related to the core part of the test;
b) Give the user a reasonable motivation for each task, for the user, the functionality of the product is not important, but the important thing is that the user wants to use the product to achieve the purpose;
c) Give the task a real scenario, after all, the user wants to use the product to achieve the goal;
c) Give the task a real scenario.
In short, the design of the test task is "who wants to do what in what situation", grasp the "people", "situation", "goal" three elements. The goal of the three elements.
The purpose of writing test scripts is to prevent confusion during formal testing when you don't know what to ask or what to record. According to the end of the test results of the statistical analysis of the dimensions of the start, combined with the task list, bottom-up analysis should be in the test need to ask questions, and the need to observe the attention of the point. For example:
Ask: Enter the page, first see what? And then what do you see? Where will the user want to click? Did the user notice the numbers above each category? ......
Observe : What did the user explore? What did users say? Positive comments? Negative comments? User's expression, are there any negative expressions such as frowning and pouting?
The hard environment required for a test lab must be perfect, for example:
a) Lab environment: whether it is a conference room, or a specialized observation lab with one-sided glass, it must be at least clean and tidy, so that the user's mood can be relaxed;
b) Test equipment: what terminal product to use, the network environment must be smooth, and the account must be registered first;
c) Recording equipment: what kind of equipment is used, and the user must have the ability to record his/her own data. >
c) Recording equipment : audio, video, screen recording, etc., record the test process can help designers conference visit the scene, fill in some of the missing notes. You can use some audio recorders, video cameras, eye-tracker, mouse track recording, QuickTime, Mobizen, Display Recorder, SCR, Magitest and other hardware and software.
Pretesting is inviting colleagues to simulate a few times before the formal implementation to find out what's wrong with the test. Walk through the formal process, including equipment debugging, interview cut-in, questioning, observation records, results analysis, and watch the audio and video recordings, from which to find the problems during the test .
Sample size: According to Nielsen and Landauer, the number of problems found in a system during usability testing and the number of people tested follow the following formula: P=N (1-(1- L ) n )
where P is the number of problems found, N is the total number of problems in the system, and L is the percentage of users that can be found by studying a single user. value is generally defined as 31%. At this point we can plot the curve of P about n as shown below. In addition, Nielsen gives the following data: Based on 83 case studies of usability consulting recently conducted by NNgroup, we can derive the relationship between the number of usability problems and the number of users recruited in the figure on the right below. We can see that recruiting 5-8 users is enough to expose most of the usability problems in a system. Testing with more users at this point would not provide a significant improvement in our insights.
Finding users: Divided into three steps, recruiting information - & gt; screening questionnaires - & gt; telephone contact, the core principle is to "let the most representative of the user to participate in", through the understanding of the user's attitudes, behaviors, goals to select, and the user profile to match the user. The user who matches the user profile.
From the day of the test to receive the user needs to maintain a friendly attitude, can make the user to put down their guard, can have a happy mood to face the test. The opening host can do some self-introduction, hot field, ease the atmosphere of the test, indicate the purpose of the test, tell the user we very much hope you can speak freely, the test is not right or wrong, do not worry about doing a bad job, and sincerely hope that we and you together to find the product problem. And if there are audio or video recordings, be sure to let the user know that there is such behavior, but the test data will be kept confidential, you can sign a confidentiality agreement.
According to the preparation of the test script for the test, at least two staff members are needed, a moderator to do the task of guiding, asking questions about the work, an observation recorder to observe the user's emotions, movements, record user behavior, ideas. The test needs to pay attention to the following points:
a) Observation: Observe: Observe the words, identify the user's emotions, thinking, frowning, hesitation, surprise, etc., through the user's emotions to identify the difficulty of the task, the user is unable to complete the task, then the task should be terminated, and soothe the user's emotions.
b) Ask: Ask a general question, specific and objective to repeat the user's behavior just now, if the user did not take the initiative to say why, you can "by the way" ask the following "why? or through the body leaning forward, gaze and other non-verbal ways to imply that the user you want to hear more content, if the user quickly and firmly stated the reason, the reason for the credibility of the higher; if the user hesitates, or difficult to say the reason, do not continue to ask.
c) Listening: Be a good listener, when the user is speaking, you can use nodding, uh-huh, eye contact and other feedback to give the user hints that we are listening carefully, you continue to speak on.
d) Encouragement: Encourage users to think out loud, boldly speak out the problems in the test, and to affirm the user's suggestions, give encouragement and praise, recognize its value. When the user encounters difficulties, do not provide help, can be appropriate encouragement.
First of all a user test is over, you can first take advantage of the memory is still fresh, simply do half an hour to organize the work, the host and the transcripts check each other. Then wait until all the tests are completed and then usability level and problem analysis. Here's a template for summarizing usability issues as I understand them:
In Don`t?Make?Me?Think Steve mentions a simple 10-cent usability test, which can be used to do a simple usability test for a product that is in early startup and needs to iterate quickly.
"Usability Engineering" in the 11 usability experience guidelines, but also recorded together:
Here let me think of a few days ago to see the twelve UX design guidelines, but also in this amenable to a little bit, lest I can not find it, the site is also doing a very good job, the site address? : Laws of UX .
[1] Usability Engineering
[2] Turning Stones into Gold: Visitor-First Web and Mobile Usability Design Secrets
[3] How to do usability testing and user research? - fries's answer - 知乎
[4] How to conduct usability testing and user research? - Demon Ye Qiu's answer - 知乎
[5] How to do usability testing and user research? - Han Zhengshuang's answer - 知乎
[6] How to do usability testing and user research? - Kelly's answer - 知乎
[7] How to conduct usability testing and user research? - 苏静婷蝙蝠小玩子的回答 - 知乎
[8] 使用研说可用性测试操作指南
[9] 可用性测试：任务评估模型与计量方式
[10] 4方面，聊聊如何做一次更高效的可用性测试
[11] 如何进行可用性测试和用户研究？ - Everyone is a product manager
[12] With super examples, analyze the application of the "seven laws of interaction design" in design