Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2010 Jun 7:4:48.
doi: 10.3389/fnhum.2010.00048. eCollection 2010.

Striatal BOLD Response Reflects the Impact of Herd Information on Financial Decisions

Christopher J Burke  1 , Philippe N Tobler, Wolfram Schultz, Michelle Baddeley
Affiliations

Striatal BOLD Response Reflects the Impact of Herd Information on Financial Decisions

Christopher J Burke et al. Front Hum Neurosci. .

Abstract

Like other species, humans are sensitive to the decisions and actions of conspecifics, which can lead to herd behavior and undesirable outcomes such as stock market bubbles and bank runs. However, how the brain processes this socially derived influence is only poorly understood. Using functional magnetic resonance imaging (fMRI), we scanned participants as they made decisions on whether to buy stocks after observing others' buying decisions. We demonstrate that activity in the ventral striatum, an area heavily implicated in reward processing, tracked the degree of influence on participants' decisions arising from the observation of other peoples' decisions. The signal did not track non-human, non-social control decisions. These findings lend weight to the notion that the ventral striatum is involved in the processing of complex social aspects of decision making and identify a possible neural basis for herd behavior.

Keywords: amygdala; anterior cingulate; decision making; herd behavior; neuroeconomics; ventral striatum.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Behavioral task. After a variable inter-trial interval (ITI) participants were presented with artificially generated stock return data for 1.5 s. Five different distributions (high mean, high variance; high mean, low variance; low mean high variance; low mean low variance; scrambled image with no discernable mean or variance) of returns were randomly presented. This was followed by social information consisting of human faces with previous decisions for the presently viewed stock. Decisions were denoted by checkmarks or x-marks above the faces. Four different social conditions were used; everyone buys (“+4”), a split group (2−2), everyone does not buy (“−4”) and a scrambled decision with no discernable information (no social, “NS”). On half the trials chimpanzee faces were presented with the same social conditions as a non-human, non-social control situation. After a jittered interval of 2–4 s the participant made a decision to “adopt” or “reject” the stock using the button box when the fixation cross was circled.
Figure 2
Figure 2
Behavioral data (n = 17). (A) Participants were sensitive to social information (one-way ANOVA, F(3,17) = 67.14, p < 0.001) and choice behavior was modulated significantly more by human social information (two-way ANOVA, F(3,17) = 9.95, p < 0.001). (B) Response time data. Participants had significantly longer response times for contrarian decisions ‘C’ than aligning decisions ‘A’ during human information trials (student's t-test, t = 1.86, p < 0.05). There was no significant difference for chimpanzee information trials (student's t-test, t = 0.84, p = 0.40). (C) Behavioral stock value preference for each participant (arranged in ascending order of mean preference). The individual stock value preference was calculated by subtracting the percentage of buy decisions for the low mean stock from the percentage of buy decisions for the high mean stock. A median split of the participants produced high and low mean-preferring groups used in subsequent mean preference contrasts. (D) Behavioral preferences for stock variance. Participants chose the high variance stocks more often than low variance stocks (student's t-test, t= 4.1, p < 0.001).
Figure 3
Figure 3
Responses to social information in the absence of stock information. (A) Behavioral changes in buy decisions for the scrambled stock as a function of human social information (one-way ANOVA, F(3,17) = 18.24, p < 0.001). (B) Activity in ventral striatum tracking social information after presentation of the scrambled stock (peak at 10, 12, −6; p < 0.05; z = 2.5). (C) Parameter estimates showing ventral striatum tracking human social information.
Figure 4
Figure 4
BOLD responses to stock information. (A) Activation in ventral striatum reflecting subjective preference for the stock mean (peak at 9, 18, −9; z = 4.2; p < 0.005). (B) Parameter estimates for the two groups. The left hand side of the chart displays the parameter estimates from the low mean-preferring group whereas the right shows the high mean-preferring group. (C) Activation in ventral striatum corresponding to stock variance (peak at 9, 12, −3; z = 3.2; p < 0.05). (D) Parameter estimates for the high–low variance contrast.
Figure 5
Figure 5
BOLD response reflecting average participant behavior. (A) Activation in ventral striatum covaried with herd influence (peak at 6, 15, −6; p < 0.05 FDR, small volume correction; z = 3.7). This corresponds to the change in behavior as a function of human social information minus the change in behavior as a function of chimp information. (B) Correlation of first-level parameter estimates from individualized herd influence contrasts and participant herding coefficients. Ventral striatum activity covaried significantly with the subjective degree of herd influence (Pearson's r = 0.544, p = 0.024; Kendall tau = 0.356, p = 0.048). (C) Comparison of second-level parameter estimates (betasGLM) for each social condition from the same analysis and region of activation shown in (A). (D) Adjusted BOLD time courses in right ventral striatum time-locked to the presentation of social information. The behavioral relationship is evident between 3 and 7 s after the onset of social information.
Figure 6
Figure 6
BOLD responses at the time of decision. (A) (Left) Activation in left amygdala (peak at 18, 0, −21; z = 3.5; p < 0.05) corresponding to an alignment decision. Alignment decisions were defined as choices that coincided with the social information of the present trial. (Right) Parameter estimates for left amygdala corresponding to alignment ‘A’ and contrarian ‘C’ decisions for human and chimp information. (B) (Left) Anterior cingulate cortex (peak at 0, 45, 15; z = 4.2; p < 0.005) was more active during contrarian decisions following human information, but showed no differential response to alignment/contrarian decisions following chimp information (right).
See this image and copyright information in PMC

References

    1. Adolphs R., Tranel D., Damasio A. R. (1998). The human amygdala in social judgment. Nature 393, 470–474 10.1038/30982 - DOI - PubMed
    1. Aharon I., Etcoff N., Ariely D., Chabris C. F., O'Connor E., Breiter H. C. (2001). Beautiful faces have variable reward value: fMRI and behavioral evidence. Neuron 32, 537–551 10.1016/S0896-6273(01)00491-3 - DOI - PubMed
    1. Ambroggi F., Ishikawa A., Fields H. L., Nicola S. M. (2008). Basolateral amygdala neurons facilitate reward-seeking behavior by exciting nucleus accumbens neurons. Neuron 59, 648–661 10.1016/j.neuron.2008.07.004 - DOI - PMC - PubMed
    1. Anderson L. R., Holt C. A. (1997). Information cascades in the laboratory. Am. Econ. Rev. 87, 847–862
    1. Baddeley M. (2005). Housing bubbles, herds and frenzies: evidence from British housing markets. CCEPP Policy Brief No. 02-05, Cambridge Centre for Economic and Public Policy, Cambridge, UK

LinkOut - more resources