Toddlers and touch screens

In spring 2012, BBC Children’s teams were preparing to launch their first mobile-specific HTML5 sites and games. Elizabeth Valentine-House writes about her research into how toddlers use mobile touchscreen devices.

In spring 2012, BBC Children’s teams were preparing to launch their first mobile-specific HTML5 sites and games. This required help from the BBC’s research and development team to gather data on how toddlers use mobile touchscreen devices. There is a general assumption that even very young children today are able to pick up and use mobile touchscreen devices with speed and ease. However, this is because touchscreens quickly teach children the link between a cause (a touch) and an effect (the thing that then happens on the tablet or smartphone). With touchscreen devices there is much less disconnection between cause and effect than with a desktop PC; there is no keyboard, mouse or track-pad interrupting the correlation between the child’s action and the result.  

Regardless, it is not safe to assume that a child using a touchscreen device can comprehend its technical infrastructure or navigation mechanisms, that they can read any text on the screen or that they have the physical strength and dexterity to interact with finesse. These are longer-term learning and growing processes, and touchscreen design for children must take this into account. The research was therefore an opportunity to explore these issues further via the prototype mobile site and games.

A research plan was developed to understand how toddlers and young children use mobile devices in their own homes. The main questions were:

  • How do they hold and interact with them?
  • What do they use them for?
  • Where and when do they use them?
  • Whose device is it?
  • How do the parents feel about their child’s use of devices?
  • How do the children respond to and use the mobile site and games?

Although testing products in a lab setting can yield scientific findings with controllable variables and limited distractions, it is also important to observe just how people use their own technologies in familiar day–to-day environments. This is especially the case with children, who can easily be overcome by shyness, excitement or “classroom mentality” in a lab, and are more confident and comfortable at home. Visiting children in their homes also reveals environmental, sibling and parental influences. Interactions between siblings and the language they use to describe experiences offers a valuable insight into children’s mental models of digital media.

Although in-home research provides a rich snap-shot of interactions, it is a hard environment to control because of its multiple distractions. There are also many practical considerations such as: CRB checks, finding note-takers to help capture the rich data, arranging travel, spacing sessions adequately, choosing recording devices carefully for use on the move and setting them up with minimum fuss/risk to people in the house. Also, when showing prototypes to children on their own devices, it is important to manage their expectations and make sure that they understand that their access to the prototypes is just for that session.

Obviously, in-home research won’t provide evidence of behaviour in other places where devices are regularly used (such as cars, trains/buses, or waiting areas/restaurants). Some children will exaggerate their use, whilst others may remain shy, be tired or generally un-talkative on that day. In these cases, researchers can check with parents what the true level of use is, modify their communication style to encourage a response, or try leaving for a while or observing from a distance.

For this particular study, eight children aged between three and five years old were visited for up to an hour in their homes, with their parent(s) and sometimes their siblings also present. A mixture of children were observed, with a deliberate spread across variables such as: gender, being left- or right-handed, family socio-economic status, familiarity with mobile devices, and the type of devices used. The devices focussed on were: iPhone 4/4S, Samsung Galaxy S2 or similar mobile (Android), Galaxy Tab Tablet or similar, and iPad 2 or later model.

The research took three stages: firstly, the child showed what they used already and how, going through their favourite things to do on the device(s). Whilst they did that, the parents would discuss their child’s use of the device, and clarify or moderate what the child was saying. After this, the child was shown the prototype mobile site and asked to find something to do on it. The mobile games were then explored with the child one by one, rotating the order where possible. Factors observed and recorded were: what the child tried to do, how they interacted with the device, and what they felt about the games.

There were five games featuring a mixture of characters; this variety was very useful in keeping the children motivated to take part. The games were:

  • Mr Tumble Dress Up
  • Paint with Little Monster
  • Mike the Knight Pairs
  • Robert’s Robot Panic
  • Fun with Justin

There were many observations from the visits. Some findings related directly to the prototypes and included suggestions for improvements, additional game ideas, interaction problems and potential misunderstandings. Other findings related to general use were grouped into themes; manner of use, current apps and sites used, and usage issues and concerns (mostly parent-led).

For example, the findings raised questions over implementation of the games, which were made to run in a browser as HTML5 games, instead of as apps. There were many business reasons for this, including maintainability and the avoidance of unfair competition in apps stores. However, the research illustrated that children are used to the apps model and understood how to enter and exit a game by clicking on a certain icon, as opposed to navigating back or forwards through browser pages. Children are not accustomed to entering in text or URLs.

Whilst web links can be saved to the home screens as icons, there are additional issues with web-based games. For instance, the children often mistakenly accessed the browser controls and navigated away from the game, causing confusion and requiring a parent to step in. They also grew impatient with download times and there were sometimes delays in sound playback. Parents mentioned that they tended to pre-load apps before going out, in order to avoid mobile internet charges and mobile internet dead-zones. Parents preferred that children used apps, as they were “walled-gardens” that children could play within, without frequently needing to ask the parent for help with navigation.

A further problem with web-based games was that in order to play these games, the device needed to be rotated and held in landscape. To indicate that this was needed, a screen was shown telling the child to turn the device around. The children could not read this, so this screen needed to be made visual. The main problem, though, was that children don’t hold devices like adults do and struggled to understand how to trigger the rotation. Also, iPads are heavy, so children would generally rest them on their lap, level or even tilting away from themselves. Since their hands are smaller, it is harder for them to grip a device in landscape and still play a game.

Expecting children to use fingers and thumbs in a refined manner was also not appropriate. Children’s presses may be very light and might not trigger an action. Children would also tend to press things that looked touchable with more pressure if they didn’t work on the first touch. In Robert Robot’s Panic, children were unsure what they needed to touch, what happened when they did do a successful touch, and what the objective of the game was; clarity of what is touchable and what the effect of doing so is necessary.

A game that worked well was Fun with Justin; where items could be dragged onto Justin resulting in him making a funny face and a humorous noise. Children delighted in this simple interaction and clear outcome. Mr Tumble Dress Up was also enjoyable, but children did not notice the green ticks to the side and were content to dress Mr Tumble up freely. They became frustrated when they could not move forwards with this game, because they did not realise that they needed to dress him according to the stated weather before moving on. From these sessions, it appears that games where strong and simple cause and effect are the focus work well, but games with subtle rules or requiring a finesse of control are not so appealing for toddlers and young children.

These findings offer a fascinating insight into how children use mobile touchscreen devices. The research impacted the design of the site and games, improving the experience for children. Actionable changes were discussed and implemented before the launch of the site and games, with further-thinking suggestions being saved for later. The full findings were documented in a technical paper and were also summarised in a presentation to the User Experience and Design (UX&D) team working on the site and games. As well as the findings, the value of in-home research with toddlers was demonstrated and the experience of this was fed back to the UX team.

"Designers need to consider children’s mental models, understanding of navigational paradigms, their need for strong correlation between cause and effect, and physical abilities to interact reliably and consistently." - Elizabeth Valentine-House

Overall, the use of such games by children has the ability to engage, teach and motivate. However, designers need to consider children’s mental models, understanding of navigational paradigms, their need for strong correlation between cause and effect, and physical abilities to interact reliably and consistently. It is important to not assume that children use and understand mobile devices the way that adults do, however dextrous they may appear.