A bilateral frontoparietal network underlies visuospatial analogical reasoning

Continuing my revival of this paper a day blog, today I'm summarizing this extremely well written paper by Christine Watson and Anjan Chatterjee at U Penn.  They investigate the difference in regions activated during analogical reasoning, and a similar control task which did not involve relational integration.  Analogy trials required participants to select the answers which contained the same pattern of spatial relations as the source, and the correct answers for item trials contained the same items as the source in any order.

Fig. 1. Stimuli from the analogy and item tasks. A) An analogy trial on which the critical relations between source and correct answer are between shapes. B) An analogy trial on which the critical relations between source and correct answer are between colors. C) An item trial on which the correct answer and foil differ only in shape. D) An item trial on which the correct answer and foil differ only in color.

This control task (the item trials) is particularly important.  A major challenge we face when investigating analogical reasoning is the fact that analogy questions tend to be harder than the control task questions.  In this case item trials were matched for accuracy and even took slightly longer to answer than the analogy trials.  The researchers covered this difference by running analyses both with and without response time as a covariate of no interest.  It made no difference to the results.

And what were the results?  In their whole brain analysis, the only found right inferior frontal gyrus (IFG) activated more for analogy than item trials.  In their region of interest analysis, they showed the same difference in activation for bilateral rostrolateral prefrontal cortex (RLPFC, an area associated with relational integration), bilateral IFG and bilateral inferior parietal cortex. (I would have liked to see middle frontal gyrus tested to see if there was a difference in working memory load.)

Fig. 3. A) Whole-brain activation map for regions that responded more strongly in the analogy task compared to the item task. Map has been thresholded at α = .05 (t = 4.22) corrected for multiple comparisons with at least 20 contiguous voxels. B) Right and left empirically-defined RLPFC ROIs for analogy > item contrast, both ps < .05. Although ROI statistics were computed on average activity, voxel-wise activation is depicted here at a level of p < .05 (uncorrected). The extent of the ROI is shown in green. C) Right and left IFG ROIs for analogy > item contrast, both ps < .05. D) Right and left inferior parietal ROIs for analogy > item contrast, both ps < .05.

A quick side note here: I'm not sure how I feel about the visualization of these results. They compute the data across entire regions of interest, but show uncorrected maps in B, C and D at a very liberal threshold of p < 0.05.  To be honest, even though I'm not super supportive of publishing uncorrected results, the paper is very clear as to you're seeing in this visualization, and their statistics are perfectly reasonable.  And it is nice to see the activation within the ROIS.

The only real criticism I have of this paper is the fact that the abstract focus on RLPFC, when really their results are most compelling in IFG and they do not demonstrate any specificity within RLPFC: it shows the same result as IFG and inferior parietal!  But, hey, there is a lot of literature supporting the hypothesis that RLPFC is involved, and they do successfully demonstrate its recruitment more on a relational integration task than an equally difficult task.  Nice work :)

A bilateral frontoparietal network underlies visuospatial analogical reasoning.

Watson CE, Chatterjee A.

Neuroimage. 2012 Feb 1;59(3):2831-8. Epub 2011 Sep 29.

PMID: 21982934