Investigating the Neural Organization of Reasoning Skills

1
Investigating the Neural Organization of
Reasoning Skills in Children and Adults Using the
Balance Scale Task M.L. Henry¹, E.A. Crone¹, J.L.
Langer², H. van der Maas³, and S.M.
Rivera¹ University of Calif., Davis¹, University
of Calif., Berkeley², University of Amsterdam³
INTRODUCTION Proportional reasoning undergoes
significant changes during childhood (Siegler,
1976). This is indexed by the well-known balance
task in which children are required to integrate
dimensions of weight and distance. Many
behavioral studies have shown that younger
children use less complex rules than adults when
solving these problems specifically, they fail
to integrate weight and distance dimensions (van
der Maas Jansen, 2003). Recent evidence
suggests that distinct stages of cognitive
development can be characterized by differential
profiles of brain activation (Rivera Henry,
2003). The goal of the current functional MRI
(fMRI) study is to test whether the protracted
development of the ability to integrate weight
and distance dimensions on the balance scale task
is associated with a prolonged maturation of
regions within PFC or interactions between PFC
and parietal cortex which have been found to be
critical in studies of other types of reasoning
abilities (e.g., Menon, Rivera et al., 2001,
Christoff et al., 2001).

• CONCLUSIONS
• Behavioral results revealed that both adults and
  children were significantly slower and
  significantly less accurate on the dual dimension
  trials (weight different, distance different)
  than on the other trials.
• Adults showed activation in the predicted regions
  of interest including left DLPFC (Christoff, et
  al., 2001), left Angular Gyrus, and left inferior
  and superior parietal gyri (Menon, Rivera, et
  al., 2000 Stanescu-Cosson et al., 2000).
• Across all regions of interest, children showed
  less activation than adults.
• We observed no significant differences in
  activation for adults between conditions. By
  contrast, children showed a marked decrease in
  activation, particularly on the dual dimension
  trials, within the left DLPFC, left Angular
  Gyrus, and left inferior and superior parietal
  gyri.
• One interpretation of these data is that
  children, who are generally not able integrate
  weight and distance dimensions on the balance
  scale task, also do not recruit relevant brain
  regions when task trials require that they
  perform this level of reasoning.
• Future analyses will include comparing
  participants on level of logico-mathematical
  reasoning, rather than by age, to further
  elucidate the brain-behavior relationships
  involved.

TASK For each trial children were asked to answer
the question What will happen when the hand is
removed? Will the scale tip down on the left,
tip down on the right or balance? Subjects
responded with a left, middle, or right manual
button press. Pictures were counterbalanced to
produce an equal number of left, right, and
balance answers and an equal number of trial
types
Weight Different, Distance Same W1D3
Equal W9D4W9D4
Weight Same, Distance Different W1D7W1D2
Weight Different, Distance Different W6D3W9D2

• Each stimulus was presented until the participant
  selected a response.
• Participants had 20 seconds to respond to each
  problem. For the remainder of the trial, a
  fixation cross was presented
• 36 trials/condition 4 runs 36 trials/run

BEHAVIORAL RESULTS
Proportion Correct
Reaction Time
condition p age group p .82
condition p age group p .10
PRE-SCAN ASSESSMENT Subjects were asked to
participate in a reasoning skills assessment
prior to scanning to obtain their level of
logicomathematical reasoning. This was a
semi-structured interview in which children were
asked to solve various problems involving the
balance scale and explain their answers. For
example, a subject was given two weights and
asked to make the scale balance. After balancing
the two weights, the subject was asked, Can you
make it balance another way? and Why does that
work? These results are currently being
analyzed.
REGION OF INTEREST ANALYSES
L Inferior Parietal
L Superior Parietal
L DLPFC
L ANG
condition p .23 age group p age group p .1
condition p .055 age group p age group p .07
condition p .056 age group p age group p
condition p .17 age group p age group p

• fMRI METHODS
• Subjects
• 5 Children (1 male, 4 females)
• Aged 9.47-10.98, mean age10.17, SD0.56
• 12 Adults (4 males, 8 females)
• Aged 19.89-24.95, mean age22.13, SD1.83
• fMRI Scanning
• 1.5T GE scanner
• Echoplanar Imaging (EPI) sequence
• 27 axial slices parallel to AC-PC (4 mm slices
  1mm gap)
• TR 2 sec, TE 40 msec, 3.125 x 3.125 mm
  inplane resolution
• Analyses
• Standard statistical analysis using SPM2,
  including
• Normalization and interpolation to 2x2x2 mm
  voxels
• Spatial smoothing (8mm Gaussian)
• Targets were modeled as events.
• Regions of Interest (ROIs) were generated from a
  fixation contrast based on the 12 adult subjects.
  Analyses were run using MarsBaR, a region of
  interest toolbox for SPM (http//marsbar.sourcefor
  ge.net)

WHOLE-BRAIN ANALYSES
Equal
Weights Different, Distance Same
REFERENCES Christoff, K, Prabhakaran, V, et
al. (2001) Rostrolateral prefrontal cortex
involvement in relational integration during
reasoning. Neuroimage, 14(5),1136-49.
Stanescu-Cosson, R. et al. (2000). Understanding
dissociations in dyscalculia a brain imaging
study of the impact of number size on the
cerebral networks for exact and approximate
calculation. Brain, 123 (11)2240-55. Menon
V, Rivera SM, et al., (2000). Dissociating
Prefrontal and Parietal Activation During
Arithmetic Processing. Neuroimage, 12, (4),
357-365. Rivera, SM and Henry, ML (2003).
Age-independent changes in brain activation
related to conservation of number in young
children. Paper presented at the Cognitive
Neuroscience Society Annual Meeting, New York,
NY, March, 2003. Siegler, R (1976). Three
aspects of cognitive development. Cognitive
Psychology. Vol 8(4), 481-520. van der Maas,
HLJ Jansen, BRJ (2003). What response times
tell of children's behavior on the balance scale
task. J Exp Child Psychol, 85(2)141-77.
Adults
Children
Adults
Children
Weights Different, Distance Different
Weights Same, Distance Different
Children
Adults
Adults
Children
ACKNOWLEDGMENTS Special thanks to our research
assistants, Yvonne Phun and Althea Crichlow, for
their assistance in the collection and analysis
of this data.