Website User Experience Analysis: Interpreting 12 misunderstandings of eye movement

Source: Internet
Author: User

The attention of the academia and the business community to the application of eye-tracking technology (eye movement, for short) is an upsurge of discussion every once in a while. With the innovation and promotion of technology, more and more users of the research industry began to introduce eye tracking technology in product design and development process, trying to read the products and experience in the eyes of users.

However, due to the lack of cognition of eye movement technology and the asymmetry of information, the misunderstanding of this research method is constantly appearing. This article will enumerate some of the user research areas of eye movement misunderstanding, and try to answer the correct application.

  1 "Make an eye move first and see what the user is looking at?"

A person who has an eye for knowledge usually throws such questions at the researcher. If you blindly take such a task, without the needs of segmentation and planning of the eye movement Test, the end of all people will be only a few beautiful browsing trajectory map and focus on the density of hot zone map.

Imagine when the software interface needs to make a big adjustment, with eye movement to record the "static design of the software" and "clickable, interactive" software demo version, the user's attention point will be how much different? When analyzing eye movement results, ambiguous test objectives can be difficult to answer the question of "should this/may be so". Usually not obvious conclusions, but also trigger a new round of additional test tasks. Before understanding and communicating the test, you should be clear about the purpose of the test and the problem that you want to answer with the data.

Each task has a clear test start, the completion of nodes, so that the results are easy to analyze and contrast, is the meaning of eye movement data.

  2 "eye movement is the hot zone map and trajectory map." ”

Nielsen discovered in 06 that people were browsing the web with the letter "F" type of reading habits. This f-type Hot zone map is cited on many occasions, but very few of the hot areas are based on the user's eye stay, the time of stay, in what time range generated? Does the user browsing habits change because of the structure and time of the website instead of the F type? People who read the data in the report often say, " I want a hot zone map of xxx, but not to consider the various constraints behind. The same misfortune will befall the "trajectory map".

Eye Moving Hot Zone map

By the number of points of concern by the length of time

This is not to blame for the people who read the report, because these two graphs often appear in the report for more intuitive understanding and presentation of the results. In fact, each project can analyze and expound the user's cognitive behavior from different aspects and angles, and every angle and measure will bring the understanding of user browsing and cognitive behavior. Therefore, it is easier for readers to grasp the problem because of the hot zone map/trajectory map with proper eye movement index.

  3 "The red dot of eye movement is the exact position that people look at." ”

Whether you're looking at a visual or a trajectory, someone will think that red dot/cross 100% is where the user sees it. Not all of this is true.

First of all, the current commercial non-contact eye movement of the data sampling rate is not as high as we think, so the display of the gaze point may be the actual point of view of the user can be biased. Moreover, when the user changes the position of the relative screen during the testing process, the camera that records the eye movement data is very likely to "miss" the user's eye movement. When the data is superimposed on the test material, the "drift" of the eye movement trajectory occurs.

(Eye movement trajectory drift: the trajectory shows that the reader is reading text, but the area does not match.) )

Second, in general, the vision of the human eye is divided into high-definition visual areas (central Fossa vision: foveal vision) and low definition visual areas (Edge Vision: Peripheral vision). The high definition area that corresponds to the screen is usually greater than the focus (fixation) displayed by the red dots.

For example, when we look at a computer screen, the visible HD area is almost two fingernails.

When the human eye is looking at a more familiar or larger object, it is still possible to notice the blurred image area (parafoveal& peripheral vision).

As shown in the following illustration, although the user does not see the camera and purse directly in the advertisement, the user is aware of the high likelihood of the camera and the purse.

Therefore, we cannot say that the user does not see anything, and can only say that they do not see anything directly up to it.

  4 "Eye movement data is easy to explain. ”

"I want to know where the user is going to look"--the phrase that is often heard will be a research goal.

The person who said it seemed to have a very clear case: if I knew what the user was looking at, I would know how to change the thing. In fact, at the end of the time, when we take the browsing order and the hot zone map that the user is looking at, we can sometimes get a hold of ourselves. In some places the hot zone map is very red, which means that users like to see this place? Or is this where the user doesn't understand what it means?

After seeing the eye movement result, the other side may ask:

"So what?" he said.

"Based on the results of eye movement, do you have any suggestions to revise?"


In fact, eye movement can only answer you "see what", but can't tell "why see".

For example: From a app horizontal, vertical version of the eye moving hot zone map distribution, horizontal version of the design of the red area more, users seem to prefer the horizontal version. It is difficult to explain it by eye-movement data alone. The impact of test content, user's personal browsing habits and other factors, may cause users to see the horizontal version more, it is possible that users in the horizontal version of the need to spend more time to extract information.

It is difficult to make a reasonable and comprehensive interpretation of the data by eye movement alone.

  5 "All usability tests can benefit from eye movement." ”

A common understanding of eye movement is that if usability tests are coupled with ocular movement, the effects and results will be more conducive to finding the problem. But from each project has the time and human resources, eye movement test in the later analysis will occupy too much cost for zoning, scene slicing analysis and so on, and the final conclusion may only be browsing trajectory and attention, more specific reasons still need in-depth access to know. In contrast, the low-cost-efficient analysis of traditional usability testing is more appropriate.

In a formal usability test, the results of eye movement data accounted for a smaller percentage of overall conclusions. Take the site test as an example, if you want to know the Web page of a feature portal users will notice, and access to more sites after users of the overall impact of the site, the indicative text is easy to understand. In this case, the eye movement can only tell you he/she has seen, but the specific understanding can not be interpreted from the eye movement data, when the value of eye movement is lower.

In contrast, eye movement is better suited to answer specific questions that cannot be explained in traditional usability testing. (such as in the confirmation of online payment before the user on this page of what tangle) if the background data to show some links and page jump rate is very high, eye movement can tell you what the user is really looking for.

  6 "Learn to operate can do eye movement test." ”

At present, all kinds of eye movement in the market are functionally similar, the final analysis of the data is not significant. It seems as long as you learn the operation of eye movement Test, from the point of view of everyone to do eye movement. However, the use of the eye movement does not mean that you can do eye movement. Effective eye movement data is the premise of the study, and this is not only two mouse can be completed.

Large to the human eye work pattern, the visual cognition and the information processing process, the research methodology, the data eye movement processing and the statistics and so on, small to the project process design, the analysis parameter, even the prompt language modification and so on need to know ahead of time. After the end of the eye movement, the overall view of the hot zone map, trajectory map and seemingly irregular eye movement data combined to discover the user's browsing habits and product problems.

  7 "Eye movement is waiting for the user to analyze the data after reading." ”

Usually in the eye movement research project Process, the user researcher observes, the record, simultaneously prevents each kind of risk in the test process (program quits unexpectedly, the computer crashes and so on) the appearance. After the test began, the researcher's work seemed to be a curiosity to look at the eyes of the user browsing the eye, moving track and so on, until the test successfully completed.

The user's eye movement is displayed in real time to fellow researcher/product related colleagues, so you will be able to think of such questions at any time in your mind:

"Oh, she/he used to look like that!"

"No wonder there is no click here ...".

If you do not effectively record in time, or do not know what to remember, you will miss the user's operating habits and reasons for making mistakes. There is no explanation when you look at eye movement data afterwards.

This type of user eye movement through the discovery and confusion, should be recorded immediately, to test the end of the user's browsing performance, ask the user's understanding and operational behavior. When necessary, play the eye movement of the track video allows users to revisit the operation of the scene, with the help of sound thinking (think alound) feedback problems. This approach often exposes the root of the problem and the real needs of the user.

All in all, don't let users leave your eye-test project with questions.

  8 "Eye movements can be analyzed by looking at the trajectory video." ”

In the rhythm of promoting agile with research, it is the eye movement value that the quick output from the test results. The quickest way to do this is to watch the user's eye movement in real time, or back to the track after the test. Some people think that do not need to do formal analysis, by watching video can also draw the conclusion of eye movement, coupled with a few hot zone map can explain the user's browsing habits.

However, our eyeballs gaze more than once per second, producing 200-300 of eye-moving data points per minute. Just by looking at the video of the eye, our memory alone cannot remember and handle such a large amount of data. Let alone a video that looks at the trajectory of the eye to analyze user habits (individual/crowd) and eye movement data. What's worse, we may have a personal preference to watch the video. Because we already know the purpose of the project research, we may overemphasize the user's browsing order, the time to stay in the area of interest, and so on.

Note: The video reflects the changes in the user's eye-gaze point, but actually produces more eye-movement data than the video's gaze point.

(Eye movement trajectory video) (visual data of eye movement trajectory)

Eye movement trajectory video is suitable for conclusion illustration, not data analysis.

  9 "All eye-movement tests can be used for a sample size." ”

"30!30 is enough." ”

We often hear people say that eye movement test does not require a lot of people, 30 sample size can be. This seems to have become the standard sample size for eye movement. In fact, the sample size of the eye movement depends on the purpose of the test and the design of the experiment, and 30 is not all applicable to all the tests.

Before deciding on the sample size of the eye-movement test, the researcher should be aware that "I have to understand the" general situation "or" compare the differences between different designs "through eye movement. If there is a different design, the hope is to create a "general situation" by eye movement: Compare the difference in the test group (within-subjects), 30 samples are enough; compare the differences between the subjects (between-subjects), when the eye movement results are susceptible to individual differences , 30 samples are clearly not enough.

If you want to match eye movement to find out the problem of product usability, 8-10 samples can already discover 70-80% problem. At this point, the 30-sample eye movement data are sufficient to illustrate usability problems.

  10 "Eye movement data analysis parameters are applicable to all cases." ”

Common eye Movements Report, the total without the user's view of the browsing track, stay time and other aspects of the comparison, it seems that the value of eye movement here on the draw. When analyzing data, we are susceptible to other case analysis angles, resulting in a final analysis more like an eye-moving data report. In addition to themselves, no one can read the user's browsing habits and the problems reflected therein. Data analysis depends on which eye-movement-related data you want to use to support usability issues, browsing differences between different designs, and more. Therefore, not using some other common analysis parameters, you can effectively explain the problem.

From the common stay time, the number of visits to the line of sight, the line of sight and so on, to the few parameters such as pupil size, eye movement velocity, scanning path and so on, the rich Eye Movement Index makes the description of the user browsing behavior dimension has the spatial and temporal dimensions. With qualitative and quantitative data analysis, it will make your eye movement results easier to understand, but also more realistic to restore the user's cognitive and operational behavior.

For example, let the user complete the "buy a ticket on the booking page" task.

Time: The time distribution of the attention of the booking page (length of gaze, number of gaze, length of first gaze, etc.);

Space: The most attractive area for users (time and space distribution in the area of interest, the first attention area, the area of sight visiting and so on), the browsing track in the process of booking;

Information retrieval efficiency: The overall number of stares (the lower the number, the higher the efficiency)

Decision efficiency: First gaze to first mouse click time consumption

Content layout: Booking related information and functional areas between the number of eye jumps, browse order

The success rate of the task, completion ratio, complete time consumption, satisfaction degree, etc.

A small example of the above attempts to illustrate that there is no set of formulated parameters for all project analysis.

What's really important is that you want to tell the audience what kind of story to use with the eye-moving data, and then choose the right data for analysis.

The independent existence of eye movement data is not persuasive.

  11 "All eye movement data can be analyzed." ”

Usually in the large sample data need to clean some of the samples do not conform to the logic/data missing, and the general small sample of eye movement test, people tend to ignore the testing process of eye movement data acquisition quality. The most common data quality problems are caused by individual differences-the deviation of the eye position during the test, the camera does not have a good record of eye movement data, personal reasons to see a long time and so on. The data of abnormal eye movement of individual users (which differs significantly from other users ' average eye movement data), and those with low sampling rate of eye movement data should also be considered for cleaning.

It is difficult to judge the cause of these "problem data" only from the numbers. Perhaps because the user's eye movement has not been captured, may be just the user's personal habits, but also may be the user in the test forgot the task at that time. We can determine whether the data should be removed from the overall sample with the playback video of the eye movement trajectory. In data analysis, these poor quality data can cause the browsing behavior to be misinterpreted.

  12 "visual results of eye movement results are reliable." ”

The Hot zone map/trajectory map has become a standard output form of eye movement data visualization, and people may even think that it is not an eye-movement analysis without seeing such a graph. The visualization of eye movement data enables people to understand the user's focus and browse order very quickly even without relevant knowledge.

Eye movement Trajectory Chart

Eye movement trajectory of free browsing of web pages looking for specific information on a Web page

Usually when we see these visualizations, we rarely take into account the circumstances in which the picture is generated. The same Web page, allowing users to browse and find a news of interest, the resulting eye movement will be completely different. The data and eye-movement results generated in the appropriate scenarios will make sense.

So the next time you see the Hot Zone map/other forms of eye movement visualization, remember to ask what is the time range and what you do when the eye movement results.

(This article originates from Tencent CDC Blog, please indicate the source when reprint)

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.