Please note I wrote this description for the comments section of the PsychFiles podcast, after I did an interview there. You can listen to the interview here.
The Neural Correlates of Visual and Verbal Cognitive Styles
David J. M. Kraemer, Lauren M. Rosenberg, and Sharon L. Thompson-Schill
J. Neurosci. 2009;29 3792-3798.
Subjects were given a verbalizer/visualizer questionnaire. They were also given several verbal and visual ability tests from the Wechsler Adult Intelligence Scale. The authors made a composite of these subtests to form a single visual ability scale and a single verbal ability scale. Based on previous work, they expected that these would correlate with the verbalizer/visualizer questionnaire. In other words, if I’m good with words, I know it, and I call myself a verbalizer, but if I’m good with spatial stuff I call myself a visualizer. They didn’t observe this expected correlation. BUT if you take a difference score of the verbal and the visual ability tests and correlate that with the score on the verbalizer/visualizer questionnaire you get a highly significant correlation—0.63. [Comment: taking a difference score is a bit odd –it’s not clear what verbal ability minus visual ability means. Taking the difference score might make sense if the two measures were opposite ends of a single dimension. Then taking the difference score would highlight any differences in ability. But in fact the verbal and visual ability measures are themselves correlated, +0.57.]
The key behavioral task involved making a similarity judgment of figures. Each figure had three dimensions (e.g., “orange, circle, stripes”) and could be shown as a verbal description or as a figure. (i.e., you would actually see a striped orange circle). Subjects would see one such stimulus (the target), followed by two probes. The subject’s task was to say which of the two probes was more similar to the target. (The more similar probe always shared two of the three features with the target. Subjects weren’t told this rule, but they received feedback and most achieved good rates of performance). The target could be presented either visually or verbally, and so could the probes.
Here’s why the task was interesting. The task is easier for everyone when the target and probes are in the same format: that is, if both are visual or both are verbal. BUT when you see (for example) a verbal target, you don’t know whether the upcoming probes will be visual or verbal. If you’re a visualizer, it would be smart to convert the verbal description to a visual image, because you’re better with visual material.
And the brain imaging data support the idea that people do just that. The more someone looked like a “verbalizer” on the questionnaire, the more likely they were to show increased activity in “verbal” parts of their brain (left Supramarginal gyrus) when they were presented with simple pictures. The more someone looked like a “visualizer” on the questionnaire, the more likely they were to show increased activity in “visual” parts of their brain (fusiform gyrus) when they were presented with words.
The authors did a number of other analyses to rule out alternative explanations (e.g., that activation differences were due to better visual and verbal abilities in the visualizers and verablizers, respectively).
COMMENT:
This study is interesting because it is the first set of data showing a neural correlation with a stated style preference. This verbalizer/visualizer distinction seems a little less interesting because it is correlated with ability. It’s not controversial that some people are better with words and some with images/space. People know what their proclivity is. What’s new in this experiment is showing that, given the chance people will translate from the less- preferred to the more preferred representation.
What was not emphasized in the paper was that subjects did not score any better on trials that matched their preferred modality than trials that did not. That is, you would expect that verbalizers would be faster and more accurate when doing the all-words trials, and visualizers would be faster and more accurate when doing the all-pictures trials. But they weren’t.
The results of this experiment might be of some interest to cognitive neuroscientists, as it indicates that people may have enduring strategies that they use on different tasks, and that they can tell you something about these strategies. There is not much here for educators, however. This is another example of a learning style distinction that does not help predict when people will find a task more or less difficult.
Back to the Learning Styles FAQ page.
The Neural Correlates of Visual and Verbal Cognitive Styles
David J. M. Kraemer, Lauren M. Rosenberg, and Sharon L. Thompson-Schill
J. Neurosci. 2009;29 3792-3798.
Subjects were given a verbalizer/visualizer questionnaire. They were also given several verbal and visual ability tests from the Wechsler Adult Intelligence Scale. The authors made a composite of these subtests to form a single visual ability scale and a single verbal ability scale. Based on previous work, they expected that these would correlate with the verbalizer/visualizer questionnaire. In other words, if I’m good with words, I know it, and I call myself a verbalizer, but if I’m good with spatial stuff I call myself a visualizer. They didn’t observe this expected correlation. BUT if you take a difference score of the verbal and the visual ability tests and correlate that with the score on the verbalizer/visualizer questionnaire you get a highly significant correlation—0.63. [Comment: taking a difference score is a bit odd –it’s not clear what verbal ability minus visual ability means. Taking the difference score might make sense if the two measures were opposite ends of a single dimension. Then taking the difference score would highlight any differences in ability. But in fact the verbal and visual ability measures are themselves correlated, +0.57.]
The key behavioral task involved making a similarity judgment of figures. Each figure had three dimensions (e.g., “orange, circle, stripes”) and could be shown as a verbal description or as a figure. (i.e., you would actually see a striped orange circle). Subjects would see one such stimulus (the target), followed by two probes. The subject’s task was to say which of the two probes was more similar to the target. (The more similar probe always shared two of the three features with the target. Subjects weren’t told this rule, but they received feedback and most achieved good rates of performance). The target could be presented either visually or verbally, and so could the probes.
Here’s why the task was interesting. The task is easier for everyone when the target and probes are in the same format: that is, if both are visual or both are verbal. BUT when you see (for example) a verbal target, you don’t know whether the upcoming probes will be visual or verbal. If you’re a visualizer, it would be smart to convert the verbal description to a visual image, because you’re better with visual material.
And the brain imaging data support the idea that people do just that. The more someone looked like a “verbalizer” on the questionnaire, the more likely they were to show increased activity in “verbal” parts of their brain (left Supramarginal gyrus) when they were presented with simple pictures. The more someone looked like a “visualizer” on the questionnaire, the more likely they were to show increased activity in “visual” parts of their brain (fusiform gyrus) when they were presented with words.
The authors did a number of other analyses to rule out alternative explanations (e.g., that activation differences were due to better visual and verbal abilities in the visualizers and verablizers, respectively).
COMMENT:
This study is interesting because it is the first set of data showing a neural correlation with a stated style preference. This verbalizer/visualizer distinction seems a little less interesting because it is correlated with ability. It’s not controversial that some people are better with words and some with images/space. People know what their proclivity is. What’s new in this experiment is showing that, given the chance people will translate from the less- preferred to the more preferred representation.
What was not emphasized in the paper was that subjects did not score any better on trials that matched their preferred modality than trials that did not. That is, you would expect that verbalizers would be faster and more accurate when doing the all-words trials, and visualizers would be faster and more accurate when doing the all-pictures trials. But they weren’t.
The results of this experiment might be of some interest to cognitive neuroscientists, as it indicates that people may have enduring strategies that they use on different tasks, and that they can tell you something about these strategies. There is not much here for educators, however. This is another example of a learning style distinction that does not help predict when people will find a task more or less difficult.
Back to the Learning Styles FAQ page.