In the last posting in this series on interaction in online media, we looked at those forms of interaction that require the user to make a selection from a number of provided options. These came in a wide variety of forms – multiple-choice questions, pictorial selections, event selections (from audio and video material), rating scales, hyperlinks and menus of every shape and size. While convenient and easy to work with, selections are limited by the fact that the user can only work with the options provided and is therefore not able to express a preference which has not been predicted in advance. And when selections are used as a basis for assessing knowledge, the user is faced not with recalling an answer from scratch (think University Challenge) but with the much simpler one of recognising the right answer from a list (think Who Wants to be a Millionaire).
There are many ways in which users might be asked to interact by supplying a response of their own making:
Textual responsesThe user may be required to type a short text string into an input field (perhaps to enter their name onto a form or make a search query) or a longer free text response into a larger text box (as with a chat program, a free-text field in a feedback form, or when responding to an essay-type question). In the case of the former, the program can more easily interpret the response, point out or correct possible errors (“Did you mean …. ?”) and act accordingly. With free text, it is much harder for a computer to make sense accurately of what the user is trying to say, so typically this type of response has to be interpreted manually, i.e. by a human.
In a self-directed learning context, textual responses are used much less frequently than they might be (and certainly much less than they were in the hey-day of instructional design, back in the 1980s). This is probably because text responses are more tricky to set up and use than, say, multiple choice questions and authoring tools no longer typically provide the functionality needed to successfully parse (make sense of) anything but the most basic text responses.
However, as long as the question is phrased in such a way as to encourage no more than a 1-2 word response, they can be used successfully to test for recall, for example:
- What is the name of the current French President?
- Which software company created the Android operating system for mobile phones?
You can constrain the user’s answer even more by using a ‘fill-in-the-blanks’ format. You may even provide a hint by showing how many letters you require and/or the first letter of the word:
- The name of the current French President is ________ _______.
- The company that created the Android operating system for mobile phones is G_____.
If your authoring tool will manage it, you can ask for multiple free text responses:
- List the applications that make up the Microsoft Office suite.
If you’re lucky enough to have the right tool, you may be able to list a range of possible right answers, perhaps some with common mis-spellings; you may also be able to look for keywords within the user’s response, provide feedback for common mistaken responses, check for or ignore the case, and so on.
The real joy of using typed responses in a self-paced lesson is that it provides the feel of a conversation between the author and user. This technique was used widely back in the 1980s but is much less prevalent today, largely because it is not considered as a possibility. We need some new exemplars of this approach to provide inspiration.
Spoken responsesAn alternative is to ask the user for a spoken response, which requires, of course, that they have access to a microphone and that their computer has sound capability. The most common application for this is during some form of live online session using instant messaging or web conferencing software. Clearly this type of interaction will normally depend on there being another human at the other end – voice recognition is improving, but as a form of interaction is usually limited to simple selections from a list, not free form responses.
Another option, within a self-directed lesson, is to have the user record a response to a question, usually within the context of a scenario. They can then listen back to their recorded response to check how well they did. This technique works well in training call centre operators.
Numeric responsesSimple numeric responses can be simply implemented using single-line text form fields as described earlier:
- In what year was the Great Fire of London?
- How many players are there in a Rugby League team?
- What percentage of the UK population dies from heart disease?
If your software allows, you can provide feedback based on how close the user gets to the right answer.
Other ways of obtaining a numeric value are using sliders or rotary controls:
- Move the slider to set manufacturing volumes for your product range.
- Move the knob to set the required volume.
Pictorial responsesAnother way of allowing users to interact in a free-form manner is by sketching, perhaps on an electronic whiteboard in a web conferencing system or in a graphics program. When you allow users to size or adjust the position of a window or icon you are also allowing a free response.
In summary, while selecting is convenient, supplying is expressive and checks for recall (or the user’s ability to copy and paste!).
See the introductory post for this series: Interaction in online media.