I have been seeing more and more reference to the Candle problem recently and I do believe there are some incomplete and ultimately incorrect conclusions being drawn from the Candle Problem. Some of the conclusions are valid, but I believe some people are becoming quite liberal with applying the laws of science to support their beliefs. In particular, I was watching Daniel Pink’s presentation that led me to consider writing a Blog post on this subject. You can find the presentation at the link below:
Daniel Pink’s Presentation on the Science of Motivation
The Candle Problem: (Wikipedia)
“The test presents the participant with the following task: how to fix a lit candle on a wall (a cork board) in a way so the candle wax won’t drip onto the table below. To do so, one may only use the following along with the candle:
- a book of matches
- a box of thumbtacks
The solution is to empty the box of thumbtacks, put the candle into the box, use the thumbtacks to nail the box (with the candle in it) to the wall, and light the candle with the match. The concept of functional fixedness predicts that the participant will only see the box as a device to hold the thumbtacks and not immediately perceive it as a separate and functional component available to be used in solving the task
Many of the people who attempted the test explored other creative, but less efficient, methods to achieve the goal. For example, some tried to tack the candle to the wall without using the thumbtack box, and others attempted to melt some of the candle’s wax and use it as an adhesive to stick the candle to the wall. Neither method works. However, if the task is presented with the tacks piled next to the box (rather than inside it), virtually all of the participants were shown to achieve the optimal solution, which is self defined.”
“Glucksberg (1962) used a 2 × 2 design manipulating whether the tacks and matches were inside or outside of their boxes and whether subjects were offered cash prizes for completing the task quickly. Subjects who were offered no prize, termed low-drive, were told “We are doing pilot work on various problems in order to decide which will be the best ones to use in an experiment we plan to do later. We would like to obtain norms on the time needed to solve.” The remaining subjects, termed high-drive, were told “Depending on how quickly you solve the problem you can win $5.00 or $20.00. The top 25% of the Ss [subjects] in your group will win $5.00 each; the best will receive $20.00. Time to solve will be the criterion used.” The empty-boxes condition was found to be easier than the filled-boxes condition: more subjects solved the problem, and those who did solve the problem solved it faster. Within the filled-boxes condition, high-drive subjects performed worse than low-drive subjects. ”
There are conclusions regarding Functional Fixedness that revolve around the solutions given depending on whether there are tacks inside the box or on a pile on the desk. These are not my primary interest.
My primary interest revolves around the following two facts and conclusion:
- People given the incentive perform better on solving the problem when the tacks are in a pile on the table beside the box.
- People given the incentive perform worse on solving the problem when the tacks are in the box on the table
The conclusion that is commonly drawn is that monetary compensation and reward and punishment only work for simple, repeatable tasks but actually have the opposite effect on complex, cognitive, and analytical tasks.
I believe there are incorrect conclusions being drawn as most conclusions focus solely on the monetary factor without considering the time constraint factor. There are actually two factors combined when providing an incentive to the subjects:
- A monetary incentive is provided
- The monetary incentive is based on how quick the problem can be solved. (as opposed to saying the best solution will be rewarded)
I would argue that the results do not prove that the monetary incentive alone is responsible for the results seen. If anything, I believe the factor of defining success based on duration was the major factor in the results seen. Rarely are complex solutions the first ones designed. In short, the experiments got what they measured. Focus on duration helps to deliver simple solutions quicker and hinders the development of complex solutions.
That said, I fundamentally believe that monetary rewards will not solely motivate today’s information workers to solve complex problems. I just don’t agree that this study proves that.