Frequently asked questions about Web Assessment

Lexplore’s Web assessment runs directly in the web browser using eye tracking via a webcam. For it to work properly and provide reliable results, make sure that your equipment meets the requirements listed in the technical requirements for web-based assessment. 

Start by clicking the question mark icon in the top-right corner during the Web assessment. There you will find training material and guidance that helps you solve common issues.

 

Also make sure that you meet the technical requirements for web-based assessment.

 

If the problem persists, you are always welcome to contact us at support@lexplore.com, and we will be happy to help.

 

Yes, they are exactly the same texts as before. The system where the student receives a randomised text at each assessment session works in the same way as assessment with an eye tracker.

 

Yes, you need to measure the screen to ensure correct calibration of the webcam. The measurement allows the system to determine where the eyes are positioned on the screen, which is a prerequisite for reliable results.

You only need to measure the screen once per browser and computer. The setting is then saved, and you can carry out multiple assessments without measuring again, as long as you use the same computer and browser.

For all reading comprehension questions, there is an answer option called Other answer. This option was previously called Incorrect answer and is used when the student’s response does not match any of the listed answer options.

Keep in mind that the student’s answer may still be correct even if it does not exactly match one of the options. For example, it may:

• contain synonyms of the correct answer,
• use a different grammatical form, or
• be a well-formulated explanation that clearly shows the student has understood and can answer the question.

For some students, it may feel sensitive to see how their answers are recorded. As the Examiner, you can adapt the assessment situation, for example by:

• tilting or turning the screen away while answers are being recorded, or
• setting expectations before the assessment and clarifying that it reflects what the student can do right now, and that several answers may be relevant.

Yes – as long as a sufficient number of valid data points are recorded. To present a result, at least 70% of the measurements must be valid, meaning that the student’s gaze has been registered on the screen.

If the student moves a lot, data quality may be affected. The system is therefore designed to only present results when there is enough data for a meaningful assessment.

In other words: if you receive a result, the system has determined that the data quality is sufficient.

After you have completed a Web assessment, the student’s results are automatically analysed immediately after the student has read the texts.

To view the results, in the same view where you conduct the assessment, click Published – View results, which appears next to the student’s name. This link will take you to the student’s results, which are stored under Insights – in the same place as the student’s other results.

You can also go directly to Insights and search for the student’s results there.

The camera-based technology from Tobii calculates a combined gaze point based on both eyes. This is the point that is displayed.

Raw webcam data naturally contains more measurement variation than data from a dedicated eye tracker. Therefore, the data is processed and filtered before being shown in the portal. What you see in the visualisation is not raw data, but a quality-assured and adjusted representation of the student’s reading.

Visualisation in the portal:
Because webcam data contains more variation, we apply an intelligent adjustment that uses information about where the lines of text are on the screen. This means that the visualisation is an approximation. In some situations – for example, if a student restarts reading in the middle of the text – it may appear as if the student reads the same line several times.

What is important, however, is this: The results for reading speed and analysis are not affected.
The model is trained to interpret camera patterns and provide the same results as eye tracker recordings of the same reading. The visualisation is a support; the actual analysis is based on the underlying model.

Yes. The Web assessment is developed to provide the same type of results as our established eye tracker method.

The model is trained and validated to replicate the results from our traditional assessment and shows an accuracy of 0.97 (where 1.0 represents perfect agreement).

This means that you can use the results in the same way in your teaching – to understand the student’s reading fluency and identify the need for support. We continuously improve and strengthen the model through further validation and data collection.

Yes, you can assess the classes where you are added as a Teacher in Lexplore.