“Failure” is probably not the right word to use here, but it’s how kids will see their initial attempts to master an interface. Since, unlike our 6-8 year olds, our 8-10s don’t look for instructions before beginning, it’s likely they’ll not get it right the first time. While a 6-year-old probably won’t try something without adequate direction, a 9-year-old will size up the interface, determine if it’s worth their time, and jump right in. This presents a huge opportunity for teaching–after the kid fails the first time.
Using confirmation and error messages to provide incremental instruction is very powerful for this age group. And most kids’ sites aren’t doing this. Since kids aren’t reading directions before starting, why not use follow-up messages to teach?
I did a research study several years ago for Pepperidge Farm’s Goldfish Fun site. The site does a great job providing up-front instructions, but it misses a teaching opportunity with its follow-up messaging. The 9 year olds I got to work with opened the games and immediately started playing, ignoring the directions presented to them. When I asked about the directions, most of the answers I got included something like “Oh, this looks easy. I’ll just start playing” or “I’m really good at these kinds of games.” When the kids didn’t perform as well as they’d liked on the games, they would get discouraged and look for help.
Kids gravitated to Goldfish Fun’s Catapult Chaos— a physics game that lets players control the angle and velocity of a virtual marble launched from a spoon to knock down a bunch of items. Kids have to try and knock down as many of the items as they can.
Catapult Chaos follows all the rules for designing for kids. Instead of a lengthy instructions screen, it presents contextual tips for mastering the game. It demonstrates how to play through animation. But the kids ignored all of this. They just wanted to play.
￼Figure 1: Kids ignored the contextual help and clicked the “Skip Tutorial Mode” button to start playing.
During the research sessions, one little boy played this game almost the entire time. In fact, the only way I was able to tear him away from the computer was to show him the American Express gift card I had for him to thank him for his participation. He kept tweaking the position of his spoon and the power of his marble launch repeatedly to see if he could achieve the Total Takedown Bonus, awarded when the player knocks down everything at once. His tweaks were mostly random, though, and his frustration increased whenever his score decreased. He clicked on the “Help” link, but the instructions there were too general to help him.
Figure 2: Catapult Chaos‘ instructions are too general to provide help on how to improve one’s score.
No messages appeared during game play, and he didn’t see any type of confirmation except for a basic “nice job!” after he finished the level. He would have been able to increase both his knowledge and his score if messages displayed periodically during the round, saying “try increasing your power meter this time!” or “what would happen if you tilted your spoon a little further down?” The goal here is not to provide a cheat, or to minimize the exploratory aspects of the experience, but rather to set kids up for success in a meaningful way.
Using follow-up messaging as a teaching tool works for adult audiences, too. Think about the last time you filled out a form online. Did you read instructions before beginning, or did you just start entering your information? If you missed a field, or entered incomplete data, did you get a general “oops, try again” message, or did you get specific instructions on what you did wrong and how to fix it it? Which did you prefer?
A team I worked with at Comcast designed an interface allowing customers with subscriptions to watch programs online. The first iteration of this system was really complicated, due to technical limitations and legal complications. Customers had to download and install two different pieces of software and log in three times. The team designed a really helpful set of instructions, but during usability testing learned that no one was reading it, rather, customers were waiting for context-based feedback to help them correct their mistakes. The team revised its approach and created a series of highly personalized “error” messages that taught customers how to use the system. People were still frustrated, but the frustration was markedly less when the messages helped them learn.