This article was first published on 7 January 2005.
The problem
A while ago, a good friend asked my advice on behalf of someone else. This someone else was having problems with a class they were teaching. She was trying to teach them how to use a spreadsheet to add up a list of numbers. They were shown the function that summed them and how the numbers were persistent (if a single number changed, a single modification was needed to recalculate the entire list), but the students persisted in using a calculator (which meant keying in the entire list from scratch). This happened despite being explicitly taught differently.
Inefficient strategies
Clearly, the students preferred to use a much more inefficient strategy when faced with a problem, and for some reason they could not adopt a more efficient one. Why should this be?
It would be easy to dismiss the users as stupid and then walk away from the problem. However, research by former colleagues of mine shows that it is in fact perfectly natural with a given set of circumstances, and it happens with intelligent people under test conditions too. It is our job as usability practitioners to understand why people make decisions that are apparently not rational.
Higher goals
The problem could have been due to the presence of a higher goal - whereas the students were being taught a more efficient strategy, when they had to focus on adding up a list of numbers, the constraints presented by this higher goal interfered with the application of the more efficient strategy.
Charman and Howes (2002) investigated just this issue. Despite training people with efficient strategies, when they were constrained by having to focus on a higher goal, these efficient strategies were adopted more slowly compared to well-known (but inefficient) methods.
Given the task of creating a simple image composed of several elements and repeating that image many times, users with higher goal constraints (which was to design the layout of a room) didn’t get the idea that it was more efficient to block copy and paste as quickly as those who had no higher goal constraint (they were told simply to copy an item the same number of times as the previous condition).
Thus it seems that giving people a complex task with a higher goal inhibits efficient computer use, even among experienced users who were comfortable using computers. However, the authors note that higher goal constraints don’t inhibit efficient strategy use in themselves. Rather they reduce users’ opportunity for developing efficient strategies.
Pragmatics
So what does this imply for usability practitioners? Complex user tasks would probably need to be structured to allow the opportunity for the development of efficient strategies.
An example could be made with the interface on a mobile telephone for sending SMS’s. The problem might be that users find it overly time-consuming to send a text message. If an investigation found that the users dissatisfaction with the phone was because of their own adoption of inefficient strategies, it might be better to restructure the task. Having a series of brief instructions presented to the user might help them restructure the sub-goals (writing the message, selecting the correct recipient) as goals in themselves. Hopefully, this would enable the user to develop more efficient strategies more quickly, thus increasing user satisfaction (and effectiveness) with the phone’s operation.
The messages would only have to be something simple like “Write your message” and “select the correct recipient” with maybe some visual indication as to their location within the higher task (each subtask would scroll “onwards” to the next subtask, giving a visual cue as to their progress).
This is a simple solution, and better ones are probably available. However, the true picture would only be ascertained with user testing.
Given the earlier discussion of the students adding lists of numbers: I would suggest simply changing the instructions to the students to make them concentrate upon the task rather than the results.
Wider implications
As a pet theory, I wonder if strategies can be related to knowledge. If this is the case, then the above research has important implications for things such as problem-based learning (PBL). This technique emphasises the use of knowledge in simulations of realistic situations rather than traditional topic-based learning. The presence of higher goal constraints (which are the bread and butter of PBL) may prevent efficient use of knowledge. When teaching people, the way in which PBL is implemented may play a massive role in its effectiveness.
However, these are just untested thoughts of mine which I may research in the very near future. It all hinges on how well the findings for strategies of computer use can be related to knowledge use.
References
Charman, S.C., and Howes, A. (2002) The effect of goal constraints of strategy generation. In W.D.Gray and C.D.Schunn (Eds.) Proceedings of the 24th Annual Conference of the Cognitive Science Society, Mahwah NJ: Erlbaum, 172-177. Retrieved from Cardiff University (pdf file, 164k) on 18 December 2004.
Why do people keep making “stupid” decisions?
The problem
A while ago, a good friend asked my advice on behalf of someone else. This someone else was having problems with a class they were teaching. She was trying to teach them how to use a spreadsheet to add up a list of numbers. They were shown the function that summed them and how the numbers were persistent (if a single number changed, a single modification was needed to recalculate the entire list), but the students persisted in using a calculator (which meant keying in the entire list from scratch). This happened despite being explicitly taught differently.
Inefficient strategies
Clearly, the students preferred to use a much more inefficient strategy when faced with a problem, and for some reason they could not adopt a more efficient one. Why should this be?
It would be easy to dismiss the users as stupid and then walk away from the problem. However, research by former colleagues of mine shows that it is in fact perfectly natural with a given set of circumstances, and it happens with intelligent people under test conditions too. It is our job as usability practitioners to understand why people make decisions that are apparently not rational.
Higher goals
The problem could have been due to the presence of a higher goal - whereas the students were being taught a more efficient strategy, when they had to focus on adding up a list of numbers, the constraints presented by this higher goal interfered with the application of the more efficient strategy.
Charman and Howes (2002) investigated just this issue. Despite training people with efficient strategies, when they were constrained by having to focus on a higher goal, these efficient strategies were adopted more slowly compared to well-known (but inefficient) methods.
Given the task of creating a simple image composed of several elements and repeating that image many times, users with higher goal constraints (which was to design the layout of a room) didn’t get the idea that it was more efficient to block copy and paste as quickly as those who had no higher goal constraint (they were told simply to copy an item the same number of times as the previous condition).
Thus it seems that giving people a complex task with a higher goal inhibits efficient computer use, even among experienced users who were comfortable using computers. However, the authors note that higher goal constraints don’t inhibit efficient strategy use in themselves. Rather they reduce users’ opportunity for developing efficient strategies.
Pragmatics
So what does this imply for usability practitioners? Complex user tasks would probably need to be structured to allow the opportunity for the development of efficient strategies.
An example could be made with the interface on a mobile telephone for sending SMS’s. The problem might be that users find it overly time-consuming to send a text message. If an investigation found that the users dissatisfaction with the phone was because of their own adoption of inefficient strategies, it might be better to restructure the task. Having a series of brief instructions presented to the user might help them restructure the sub-goals (writing the message, selecting the correct recipient) as goals in themselves. Hopefully, this would enable the user to develop more efficient strategies more quickly, thus increasing user satisfaction (and effectiveness) with the phone’s operation.
The messages would only have to be something simple like “Write your message” and “select the correct recipient” with maybe some visual indication as to their location within the higher task (each subtask would scroll “onwards” to the next subtask, giving a visual cue as to their progress).
This is a simple solution, and better ones are probably available. However, the true picture would only be ascertained with user testing.
Given the earlier discussion of the students adding lists of numbers: I would suggest simply changing the instructions to the students to make them concentrate upon the task rather than the results.
Wider implications
As a pet theory, I wonder if strategies can be related to knowledge. If this is the case, then the above research has important implications for things such as problem-based learning (PBL). This technique emphasises the use of knowledge in simulations of realistic situations rather than traditional topic-based learning. The presence of higher goal constraints (which are the bread and butter of PBL) may prevent efficient use of knowledge. When teaching people, the way in which PBL is implemented may play a massive role in its effectiveness.
However, these are just untested thoughts of mine which I may research in the very near future. It all hinges on how well the findings for strategies of computer use can be related to knowledge use.
References
Charman, S.C., and Howes, A. (2002) The effect of goal constraints of strategy generation. In W.D.Gray and C.D.Schunn (Eds.) Proceedings of the 24th Annual Conference of the Cognitive Science Society, Mahwah NJ: Erlbaum, 172-177. Retrieved from Cardiff University (pdf file, 164k) on 18 December 2004.
No comments:
Post a Comment