Thanks to Matt Kmiecik, Krishna Bharani, Rob Palumbo and Riley Sicca for technical support and the Johnson Scholars Program (EG, SZ. AS), and the Psi Chi / APS Summer Research
Program (EG, RGM) for their generous support.
!Analogical reasoning has long been considered a core component of
uid intelligence (Spearman, 1927; Snow et al. 1984).
!Bunge et al. (2009) developed a visual analogy task to identify areas
in rostral lateral prefrontal cortex (RLPFC) associated with relational
integration, one core function necessary for adult analogical
!Nikitin and Morrison (2011) adapted this method for use with EEG/
ERP and identied a anterior response-locked late positive complex
ERP that indexed relational integration.
!While most developmental studies of analogy focus on young
children (e.g., Morrison, Doumas, Richland, 2011), Dumontheil et al.
(2010) demonstrated that the way the brain processes analogy
continues to change throughout adolescence.
!The present study used EEG/ERP with a visual analogy task in 13 to 25
year old females to determine whether the neural correlates of
relational integration may be partly responsible for changes in uid
intelligence during adolescence and young adulthood.
! 29 young women from age 13 to 25 participated in a multi-part
testing session where they completed the NIH Toolbox (Gershon et al.,
2010) and a visual analogy task (Bunge et al., 2009; Nikitin &
Morrison, 2011) with simultaneous EEG recording.
! A standardized uid intelligence sub-score was calculated from the
NIH Toolbox including:
! Dimensional Change Card Sort (Frye, Zelazo, & Palfai, 1995)
!Eriksen anker task (Eriksen & Eriksen, 1974) / Attention Network
Test (Rueda et al., 2004)
! The Picture Sequence Memory Test – visual episodic memory test
! List Sorting Working Memory Test – auditory/visual working
! Participants then completed 180 trials of the visual analogy task
(approximately 35min) while we monitored their EEG. Trials were
divided into 9 blocks of 30 trials each (3 blocks of each of the three
tasks, see Figure below)
Anterior Prefrontal Cortex and Relational Reasoning during
Adolescence: An Event-Related Potential Study
Robert G. Morrison Valerie Flores Elise Gagnon Sarah Zaza Amanda Sweis
Bunge, Helskog, Wendelken (2009); Nikitin & Morrison (2011)
answered a total
of 180 problems
(60 per task, half
valid and half
pressing one of
two buttons on a
-2000ms 200 ms
-600 to -200 ms
• Participants were segmented into two groups based on age for purposes of data visualization.
• There were no group diﬀerences in feature or relation task accuracy or RT between groups.
• Stimulus- and response- locked grand average ERPs from a bilateral cluster of anterior prefrontal
electrodes for feature and felation tasks for both groups are shown above.
• The neural correlate of relational integration (RI ERP) was calculated by subtracting the mean
amplitude for a group of anterior prefrontal electrodes from -600 to -200ms before the response for
feature trials from relation trials.
• Dimentional analysis below.
! While analogy accuracy and RT were not correlated with age in this sample, uid intelligence and the
relational integration ERP were.
! Analogy accuracy and the relational integration ERP were strongly correlated with uid intelligence.
! The relationship between age and uid intelligence was signicantly mediated by the relational
integration ERP, suggesting that engagement of anterior prefrontal cortex associated with relational
integration may be a critical factor in the development of analogy during adolescence.
(13-19 yo; n=13)
(19 -25 yo; n=16)
RT RI ERP
Age .40 (.03) .05 (ns) .02 (ns) .40 (.03)
Intelligence .40 (.03) .56 (.001) -.13 (ns) .48 (.008)
Accuracy .05 (ns) .56 (.001) .26 (.18) .44 (.02)
Analogy RT .02 (ns) -.13 (ns) .26 (.18) .12 (ns)
RI ERP .40 (.03) .48 (.008) .44 (.02) .12 (ns)
Mediation model with bootstrapping (Hayes, 2012)
Age Fluid Intelligence