Why Online School Education Fails (and How Good User Experience Could Save It).

After the lockdown caused by #COVID-19 most of the schools went online, and online platforms for distant learning immediately got into high demand. However, many of them appeared to be not adapted for the needs of younger learners, especially kids of the age groups of 3-5 and 6-9 years old. As a result, homeschooling turned out to be a nightmare of many parents around the world, while the usability of children E-learning became a trendy but quite painful topic for UX professionals.

While the amount of apps, online platforms, and occasional blog content for homeschooling is quite impressive, very often they do not meet the main heuristics for children e-learning systems. The information on these heuristic is quite sporadic and disseminated, let's finally put everything together.

The heuristics framework for evaluating children e-learning systems

The e-learning systems that are used by schools and colleges have core specificity of usage that determines the evaluative framework:

• the system is used on the day-to-day basis, probably for several hours;
• the system is built for several user groups, including users in early and middle childhood groups;
• the experience of using the platform is directly connected with the experience of learning itself and the satisfaction that children get from the process of schooling.

The users

The main users of the platform are kids from the age of 5 to 18, their parents, and teachers. The current heuristic framework will be centred around kids/parents-facing part of the platform, however, could also be used for a teacher-facing part of the system.

Since most of the e-learning platforms and apps are promoted as a universal solution for schools, another sensitive group of users, kids with disabilities, should also be taken into account.

The heuristics

For an online e-learning platform designed for younger children (5 and older), four main groups of heuristics can be devised:

General Usability (GU)

General usability heuristics follow the classic 10 usability heuristics for user interface design by Nielsen – Norman Group:

Visibility of system status (GH_1), Match between system and the real world (GH_2), User control and freedom (GU_3), Consistency and standards (GU_4), Error prevention (GU_5), Recognition rather than recall (GU_6), Flexibility and efficiency of use (GU_7), Aesthetic and minimalist design (GU_8), Help recognize, diagnose, and recover from errors (GU_9), Help and documentation (GU_10).

Usability for Younger Kids (UK)

The design of software for kids has been a trendy topic recently; thousands of articles have been already written. Rich starting points and further references can be found, for example, in the article on usability for young people by NN Group, the paper by Alsumait and Al-Osaimi (2010), or the article by Debra Gelman on the designing for kids.

The fundamental principles of designing for kids are, however, quite simple and fall into two main groups:

Accordingly, the following fundamentals help to support children user experience:

Child-specific accessibility (UK_1), Challenging and motivating environment (UK_2), Frictionless and attractive UI (UK_3).

Thus, Child-specific accessibility includes:

• readability (usually non-decorative fonts of no less than 14 size are advised for younger kids).
• alternative media formats are provided, especially for users with disabilities (videos are followed with captions, plain text is enhanced with voice-overs);
• there is a consistency between the motor effort that is required by the system and child's age: the system designed for young kids relies on simple user actions that do not require sophisticated mouse/touchpad/keyboard control since 6-8 years old kids are capable to perform only simple clicks and keystrokes (dragging and scrolling should be avoided);
• input/output devices are suitable for the users age groups, while the users have the necessary computer skills to use the system;
• input instances that do not have any functionality are disabled to prevent user errors;

Challenging and motivating environment presupposes that:

• the system proposes varying activities with adjustable difficulty levels; the greater challenge is presented only after a certain level of mastery is achieved;
• the principles of "conflict play" are used, creating a challenging interaction that makes the accomplishment more significant;
• the system satisfies children need for instant gratification;
• the system supports children's cognitive curiosity and encourage moving forward and progress rather than quitting.

Frictionless and attractive UI:

• skeuomorphic design is preferred: the system mimics the real-world of its counterparts as much as possible;
• the UI creates a consistent experience on all the levels (interactions, feedback, behavioral patterns, navigation, etc);
• the system proposes autocorrection and completion where it doesn't contradict the main task;
• the UI presents information in visually appealing, uncluttered, and memorable way;
• the navigation system is clear and relies on spatial and real-life metaphors;

Usability for E-Learning (UEL)

These heuristics should combine the general usability issues and the pedagogical efficiency:

• learner satisfaction with the learning content.
• learner perceptions about the applicability of the content.
• learner enjoyment of the learning experience.
• actual learning, measured via tests.

Regarding the learning experience of younger kids, the following heuristics can be used:

Engaging multimodal interactive learning materials (UEL_1), Immediate assessment and feedback mechanisms (UEL_2), Motivation to learn (UEL_3).

Accessibility

Accessibility issues, according to WCAG 2.1. criteria, usually falls into four groups of success criteria:

• perceivable information and UI;
• operable interface components and navigation;
• understandable information and the operation of the user interface;
• robust enough content to be perceived by different users.

While all the criteria are being essential to be satisfied per se, some of them appear to be essential for the user goals success. Thus, accessibility here plays also the role of assistive technologies, since young kids are quite vulnerable users that often needs assistance in coping with the content and UI elements to complete the user tasks successfully. At the same time, the system should also take into the account the diversity of kids' physiological and mental background and be ready to accommodate the user groups with vision, moving, thinking, remembering, learning, hearing, cognitive, and other impairments.

Thus, the most important accessibility criteria may be formulated as following:

• Alternative ways of presenting the content (i.e. voice-overs for tasks, video transcripts);
• Allow usage on different devices (specifically touchscreens);
• Allow individually customized critical fragments of UI (font size, dark mode, etc).

The case-study

Let's get a little bit more deep in some of the above mentioned criteria. I'm not going to present here the whole analytical report on one particular e-learning system – rather to illustrate with the help of the two systems (Eduka Klase and IXL) the most frequent issues that occur in e-learning systems for younger kids.

Quite often, the requirement of Visibility of System Status (GH_1) is not being met.

Thus, the heuristic presupposes that (see Alsumait and Al-Osaimi 2010):

a) The e-learning program keeps the child informed about what is happening through appropriate feedback within a reasonable time.

b) The child gets frequent, clear feedback that encourages him/her to carry on.

c) The child is always able to identify his/her score/status and goal in the program.

d) The child understands all terminology used in the program.

e) The child knows where (s)he is at all times, how (s)he got there, and how to get back to the main page. 

System status may refer to the two crucial levels in the e-learning system:

• external level: information about the new tasks, lesson packs, or grades;
• internal level: the way the elements of tasks are organised.

Thus, Eduka doesn't provide the visibility of system status on both levels.

External:

1. Although the parents provide their email during the registration, the system doesn't inform them when a new task appears, when the deadline for a task is coming, when a completed task has been evaluated, or about any changes in the system overall; there is also no internal system of notifications or message board;
2. My groups (Mano grupes) link doesn't work.

Internal:

1. The tasks of each exercise are graded in bulk only after all the tasks have been done. The system doesn't provide any feedback after each individual task has been done, allowing the kid to move forward even if a mistake was made. The possibility to check the mistakes is given only after the whole exercise is performed, when kids are quite unmotivated to go through the whole set of tasks again to find a mistake. This makes the process of unguided search for a mistake impossible – especially younger kids will need the guidance of an adult to go through the whole exercise again to find a mistake, while elder kids may simply abandon the exercise since a grade has already been assigned.

In this case, Immediate assessment and feedback mechanisms (UEL_2) heuristic is not satisfied. The program doesn't give any corrective directions can be realised in the form of the mistake explanation and, possibly, several training tasks to master it. A different approach is taken in IXL system where mistakes are followed by their explanation, additional practice materials, and encouragements to move further.

1. At the same time, if a program seeks learning objectives (and not only assessment ones) new attempts for the tasks should be allowed; an opposite "zero-attempts" approach doesn't encourage a kid to understand, analyse, and correct the mistake – the process which is crucial for the younger learners. The educational model behind this binary opposition of "wrong/correct" has been proven corrupted and completely inappropriate long ago but is still prospering in many countries.
2. In Eduka Klase, however, the task is locked if the deadline passes. Setting aside the same corrupted logic of an approach where the fact of "being done" is more important than the learning process itself, this algorithm doesn't allow the kids, that for different reasons could be out of the educational process for a while (illness, technical problems, etc), to accomplish the tasks they have missed.
3. The kid is not aware of his current success and is not able to identify his current status in the whole exercise or individual tasks, that also contradicts the heuristics of Challenging and motivating environment (UK_2). Neither (s)he gets any encouragements to proceed with the tasks, and needs external motivation from a teacher or parent. The way the feedback is provided is non-engaging and almost impossible to comprehend for younger kids;

Compare the feedback that is provided by Eduka platform to the feedback page on IXL for the same user age group:

The feedback page may also serve as an encouragement for continuing learning by proposing to master the topics where mistakes were detected or to move to the next topic, if a specified timeframe allows it. Thus, it corresponds to Motivation to Learn heuristics (UEL_3) that requires a program to maintain the motivation of learners and stimulate further inquiry in different ways.

Thus, a feedback page should contain not only a bright and engaging evaluation of success, as well as the encouragement to proceed, but also an easily understandable CAT (call-to-action) in order to initiate an immediate action or make sure that the user will come back later.

The dialogue below uses an encouragement ("Get more awards") combined with a call to action linked to the next task ("Continue practicing").

All these features are absent in Eduka, that isn't a learning platform, in fact. It's more an online board to check the knowledge and record an evaluation, since learning activity is absent per se.

At the same time the parents/teachers are not able to analyse the progress of the kid since:

• no information on time on task is provided;
• no attempts is allowed, so no "number of attempts – time on task – number of correct answer – subjective/objective difficulty of a task" ratio is possible, and the progress against proficiency mastery score cannot be measured.

Much more effective approach is an algorithm "task + evaluation + [another attempt, if necessary] + encouragement" where the whole process is gamified and motivates a kid for learning from mistakes, while adults get a real-time evaluation of the kid's progress and possible gaps to work on.

Moreover, such approach corresponds to the above mentioned Immediate assessment and feedback mechanisms (UEL_2) heuristics that requires the program to provide the instructor / parent with child evaluation and tracking reports. 

Below is just one example how a similar logic is realized at IXL Learning platform.

Many systems also fail to ensure the Match Between System and the Real World (GH_2), that requires the following:

a) The interface employs simple words, phrases and concepts familiar to the child.

b) It makes information appear in a natural and logical order.

c) All learning objects and images are recognizable and understandable to the child, and speak to their function.

d) The program holds to good principles of child information processing. 

e) Icons and other screen elements are intuitive and self-explanatory.

1. Thus, the program should not create any visual or verbal/logical interferences that could potentially cause frustrating mistakes. However, Eduka Klase does it – and quite frequently. The task below asks a kid to choose one of the color blocks. Providing the answers on the backgrounds of completely different colors creates a famous Stroop effect – and it's a very profound accessibility fault causing many user errors.
2. In the math task for the 1st grade the program requires the name to be written from a capital letter – otherwise recognising it as a mistake. Should a first grader really be able to enter a capital letter with a keyboard in the task on diagrams?
3. The interface rarely employs icons and is obviously verbal-centered, although the target user group is kids whose perception is still image-centered.

The heuristics of User Control and Freedom (GH_2) are usually not met together with general problems of accessibility. Thus,

a) The program should allow the child to recover from his input errors. It distinguishes between input errors and cognitive errors, allowing easy recovery from the former always, and from the latter when it is pedagogically appropriate.

b) Exit signs are visible. The child may leave an unwanted state without having to go through an extended dialogue.

c) Navigation objects and tools are kept in particular and clearly-defined positions.

d) The child should perceive a sense of control and impact on the e-learning application. He is allowed to move around in the program in an unambiguous manner, including the ability to go back to the home page or go back to previous visited sections.

e) The child can easily turn the application on and off, and can save his user profile in different states. 

Thus, Eduka Klase doesn't provide any explanation if an error occurs, nor it navigates a user to the possible solution.

The program doesn't prevent a user from a wrong format of data entry: in the math task, described above, the program allows to enter a (potentially wrong) format of answer containing a name that doesn't start with a capital letter.

This also doesn't comply with the requirement to disable impossible and/or not making sense input instances to prevent user errors and frustration. However, the program continuously fails to meet this criteria. Thus, in the task to make up the sentence, the system allows cross-matching within one category that is meaningless.

It is also crucial, especially when it comes to younger users, to ensure that navigation logics is clear and different system states are distinct. However, in Eduka hover and active state of menu is visually the same that makes it difficult for kids to orient themselves while navigating the system.

Consistency and Standards (GH_4)

a) The child experiences the user interface as consistent (in control, color, typography, and dialog design). For example; all menu instructions, prompts, and error messages appear in the same place on each screen. However, the learning objects are varied.

b) Control keys are intuitive, convenient, consistent, and follow standard conventions.

c) The e-learning program is consistent in its use of different words, situations, or actions, and it follows the general software and platform standards. 

One of the biggest platform problems is the input method that is completely not adapted to the younger kids and fails to meet accessibility criteria. The central solution that is employed most frequently in Eduka is text fields with containers for manual input, while

• helpers and/or prompt text are almost never used;
• no tooltips is provided;

Quite often text fields that look identical are used for inputting both single numbers and words that makes it difficult for a kid to understand what kind of input is required in different situations. Different tex field size, color coding, or helpers would assist in entering the right kind of data.

Kids often forget to place the cursor into the text field. If the latter would indicate that they are interactive (by increasing their salience with a color change or increase in line thickness on hover) it would help kids to remember first putting the cursor into the field and only then start entering.

The inputs that require pressing simultaneously two buttons on the keyboard should be avoided as much as possible – especially in the tasks for younger kids. For example, a task with more/less/equal signs could be designed with buttons instead of text fields that would reduce an error rate considerably.

Error Prevention and NUH_9. Help Users Recognize, Diagnose, and Recover from Errors (GH_5)

a) The program is carefully designed to prevent common problems from occurring in the first place.

b) It does not allow the child to make irreversible errors.

c) The program expresses error messages in simple language, precisely indicates the problem, and in a friendly way suggests a solution that a child can handle. 

d) The program is designed to provide a second chance when unexpected input is received (e.g., You have selected “......” in response to the question. Did you mean “.....”?) 

Thus, e-learning programs should employ some instruments to avoid the wrong by mistake entry: It is also quite hard to recover from an error because there is no clear, consistent path on how a previous action could be cancelled or a user could return one step back.

For example, in this Eduka-based task the 1st graders are required to connect the word with a letter it starts with. Once you have connected the two dots, it is unclear how you can cancel the action; after several experiments, one may discover that you should double click both of the dots, which is far from intuitive.

Recognition Rather than Recall and NUH_10. Help and Documentation Explanation (GH_6)

a) The program makes objects, actions, and options visible so that the child does not have to remember information from one part of the program to another.

b) Instructions for the use of the program are visible or easily retrievable, so that the child does not have to memorize unnecessary things.

c) The program should contain an onboarding mechanism (tutorials, test lessons) that would help the kid to get acquainted with the main functionalities.

d) The kid should be given help while using the program so as not to get stuck or have to rely on a instructor help. Help should be easy to search. Any help provided is focused on the child’s task, and lists simple concrete steps to be carried out.

e) The kid does not need to use a manual to use the application. However, (s)he has the option to receive additional guidance, instruction, or other forms of assistance as needed. 

One of the most surprising things in Eduka is that it doesn't have any user manual, not even mentioning any onboarding process. The only document that seems to be available from the main navigational page is "a user guide" that appears to be a one-page pdf document that primarily discusses the importance of using a capital letter in the tasks.

The question of accessibility is one of the most painful for e-learning platforms.

On very rear occasions the systems employ the voice-overs for verbal fragments (where it doesn't contradict the task); however, this functionality is realised in IXL.

At the same time in both programs under consideration there is no intuitive build-in functionality to enlarge the font that becomes crucial in the reading tasks. In the example below, it was difficult for a 7yo kid to read the text without being enlarged to 150%.

Not all the systems allow usage on different devices (specifically touchscreens); many require the use of specific browsers that may be problematic for parents with low computer literacy.

Thus, Eduka Klase is not compatible with Safari on iPad since it doesn't employ responsive design – in the example below half of the image together with the task itself is hidden:

This assessment can be a mile long, since the amount of UX and UI issues of different severity levels in kids e-learning software is impressive. However, simply compliance with the main groups of heuristics for such a software, together with regular user research and ongoing behavioral analytics, may significantly enhance the user experience of younger kids and give school online education the chance it desperately needs.