Ever find yourself reading a report or looking at a research result and saying, "That just can't be right?"
That's how I reacted when I heard about and then examined the Stanford-Poynter Project, which attempts to use eye tracking equipment to measure whether users gather information from text or graphics on news websites. Their preliminary conclusion: "Text seems favored over artwork for front-page attention, a preliminary analysis of an eyetracking study appears to show."
Who am I to argue with Ph.D.s from Stanford, you might ask? Well, how about trusting my intuitive problem solving ability. If you think that's too weak a basis for questioning "scientific" test results, consider the research by Peters, J.T., Hammond, K.R., and Summers, D.A., Intuitive vs. Analytic Thinking, Organizational Behavior and Human Performance 12:125-131 (1974). In experiments asking subjects to predict how two forces would affect the movement of an object, the authors presented evidence showing that
subjects who were given only perceptual cues . . . behaved in a manner . . . characteristic of intuitive thinking; i.e., a preponderance of approximately correct responses. In contrast, subjects who were given only the numerical cue values . . . behaved in a manner . . . characteristic of analytic thinking; i.e., a preponderance of precisely correct responses, accompanied by occasional extreme errors. (pp. 128-129).
The results of their experiment showed that intuitive thinking produced nearly twice as many errors. But it also showed that those errors were rarely very far off the mark, with no “extreme errors.”
The numbers-based, analytical thinking produced the exactly right answer most of the time. But it also produced “occasional bizarre responses which result from measurement and calculation errors.” (p. 126) We might add here that such bizarre responses could also result from using erroneous underlying assumptions in the analysis.
Henry Mintzberg, The Rise and Fall of Strategic Planning (1994), noted these results and suggested that intuitive thinking avoids these extreme errors because “a bizarre answer ‘feels’ wrong and is reconsidered.” To those relying on analytical thinking, however, the answer given by the numbers, however bizarre it may be, “just doesn’t feel this way, so mistakes are more inclined to go undetected.”
Essentially, purely analytic thinkers tend to trust the numbers too much, fail to reconsider unexpected results, and thus when they make mistakes, they occasionally make large ones. The third segment of test subjects in the Peters, et al. (1974) experiments, supports this proposition. Those subjects were given both the perceptual and the numerical cues and, thus, were equipped to solve the test problem using both analytic and intuitive thinking. The results matched those of the intuitive subjects almost exactly: most answers were exactly correct or nearly so, with no “extreme errors.”
While the experiment does not appear to have recorded the details of how individual subjects actually solved the problem, the results suggest that having perceptual information (in this experiment, graphic representation of the location of dots) to support intuitive thinking helped subjects “check” their work and eliminated “bizarre” results.
So, as we used to say in the '60s, "Question authority!" And trust your intuition.
Comments