Balloon Analogue Risk Task (BART)

     The Balloon Analogue Risk Task (BART) is a computerized measure of risk taking behavior. The BART models real-world risk behavior through the conceptual frame of balancing the potential for reward versus loss. In the task, the participant is presented with a balloon and offered the chance to earn money by pumping the balloon up by clicking a button. Each click causes the balloon to incrementally inflate and money to be added to a counter up until some threshold, at which point the balloon is over inflated and explodes. Thus, each pump confers greater risk, but also greater potential reward. If the participant chooses to cash-out prior to the balloon exploding then they collect the money earned for that trail, but if balloon explodes earnings for that trial are lost. Participants are not informed about the balloons breakpoints; the absence of this information allows for testing both participants' initial responses to the task and changes in responding as they gain experience with the task contingencies. Risk taking is a related, but phenomenologically distinct process from impulsivity.

     BART instrument reference:
     Evaluation of a behavioral measure of risk taking:
     the Balloon Analogue Risk Task (BART).
     Lejuez CW, Read JP, Kahler CW, Richards JB, Ramsey SE, Stuart GL, Strong DR, Brown RA (2002)
     Journal of Experimental Psychology: Applied, 8, 75-84.  PubMed ID 12075692


Scoring the Balloon Analogue Risk Task

      The primary score used to measure BART performance is the adjusted average number of pumps on unexploded balloons, with higher scores indicative of greater risk-taking propensity (Bornovalova et al. 2005; Lejuez et al. 2002). In the original version of the task, each pump was worth $.05 and there were 30 total balloons across a range of different contingencies. Balloon breakpoints range from 1-8, 1-32, or 1-128 pumps. For example, for a 1-128 breakpoint, the probability that a balloon would explode on the first pump is 1/128. If the balloon did not explode on the first pump, the probability that the balloon would explode on the second pump is 1/127, and so no. According to this algorithm, the average break point for the 1-128 breakpoint would be the midpoint of the range, or 64 pumps.

 


Methodological Considerations

      Number of trials.  Study design often limits the administration time for any single task and there is interest in knowing the most appropriate task duration that maximizes quality of data while minimizing participant burden. Studies have varied the number of balloon trials, with the most common duration being 30 trials. In studies using 30 balloons, the score across all balloons is typically more reliable than any single 10 balloon block. However under time constraints, using fewer balloons may be necessary. Research indicates that correlations with the total score are acceptable for the first 10 balloons (~.6) and are good for balloons 11-20 (~.8) with little change for balloons 21-30 (~.8). Thus, an argument can be made for any choice between 10 and 30 balloons to balance time research testing time constrains.
     Average response rates for the BART tend to be less risky than would produce maximal earnings. Participants typically exhibit adjusted scores between 26 and 35 pumps with few scoring at or above the optimal number of pumps (i.e. 64) to maximize earnings on the task. This raises the concern the most risky individuals also are making the most overall profit on the task and it is unclear why this low level of pumping occurs. One hypothesis is that the low pumping behavior might be the result of insufficient experience with the BART, but mathematical modeling suggests that adding additional trials beyond 30 balloons result in little change in pumping rates (Wallsten et al. 2005).
     Modeling behavior in a clinically diagnostic sequential risk-taking task.
     Wallsten TS, Pleskac TJ, Lejuez CW. (2005)
     Psychological Review, 112, 862-880.  PubMed ID 16262471

    
     Reward density and category.  
The BART allows for manipulation of the magnitude of reward assigned to each pump. This may be useful for studies of reward sensitivity. For instance, risk-taking declines with increasing reward/loss value (1, 5, and 25 cents per pump), and this effect is specific to those with low trait impulsivity. The highly impulsive individuals had did not adjust risk-taking in response increases in reward/loss value. This is one example of how it is important for researchers to consider the value of reinforcers when interpreting outcomes of risk taking and personality. 
BART Adults
     Measuring Risking Taking among Children and Adolescents.   A youth version of the BART (BART-Y) has also been developed which uses a point meter rather than a money counter to reinforce behavior. Participants are instructed that points may be exchanged for prizes at the end of the testing session. Reliability and validity data on the BART-Y yielded similar findings to the original BART, suggesting that dissemination of prizes as opposed to cash is a developmentally appropriate reward. 

     The BART-Y instrument reference is: 
     Reliability and validity of the youth version of the Balloon Analogue Risk Task (BART-Y)
     in the assessment of risk-taking behavior among inner-city adolescents.
     Lejuez CW, Aklin W, Daughters S, Zvolensky M, Kahler C, Gwadz M (2007)
     Journal of Clinical Child and Adolescent Psychology, 36,106-111.  PubMed ID 17206886



Psychometric Properties of the Balloon Analogue Risk Task

    
     Reliability.   Reliability of the BART has been established across a range of samples and testing conditions.
Within Session Reliability. Split-third reliability has been examined (i.e. comparing scores across the first block, middle, and third blocks of 10 balloons each), which indicate strong reliability (> 0.7). Adults tend to demonstrate modest increases in risk-taking across the blocks (typically between one and three pumps), while adolescents tend to reduce risk-taking responding from the first to the third blocks.

     Test-Retest Reliability.  Analyses of performance across sessions indicates a reasonably robust test-retest correlation (T1/T2 r = .77). Like the within session data, there is a modest yet significant increases in risk-taking with repeated administrations within a single day (T2–T1Δ = 2.2 adjusted average pumps, T3–T1Δ = 2.3 adjusted average pumps) although these are increases are smaller between testing days (two-week between T2–T1Δ = 1.2 adjusted average pumps).
     Citations for Reliability Studies:
     Test-retest characteristics of the Balloon Analogue Risk Task (BART).
     White TL, Lejuez CW, de Wit H. (2008)
     Experimental and Clinical Psychopharmacology, 16, 565-570.  PubMed ID 19086777

     Evaluation of a behavioral measure of risk taking: the Balloon Analogue Risk Task (BART).
     Lejuez CW, Read JP, Kahler CW, Richards JB, Ramsey SE, Stuart GL, Strong DR, Brown RA (2002)
     Journal of Experimental Psychology: Applied, 8, 75-84.  PubMed ID 12075692

    

     Criterion Validity.  Within adolescent studies, BART score is related to self-reported engagement in real-work risk-taking behaviors including substance use and delinquency/safety behaviors from middle adolescents to young adults in community samples as well as studies comparing clinical and nonclinical samples. Additionally, BART-Y score was shown to vary with alcohol use across three years with early to middle adolescents. Thus, BART performance appears to be related to early engagement in substance use as well as to risk behaviors related to substance use. Consideration of moderators in the relationship between BART score and risk behavior may also be important. In one study, BART-Y score was prospectively related to number of risk-taking behaviors in the past year, but only for adolescents who were low in the ability to tolerate distress. Importantly this finding persisted after controlling for gender and sensation seeking.
     Citations for Validity Studies:
     Evaluation of behavioral measures of risk taking propensity with inner city adolescents.
     Aklin WM, Lejuez CW, Zvolensky MJ, Kahler CW, Gwadz M (2005)
     Behavior Research and Therapy, 43, 215-228.  PubMed ID 15629751

     Risk-taking but not response inhibition or delay discounting predict alcohol consumption
     in social drinkers.

     Fernie G, Cole JC, Goudie AJ, Field M (2010)
     Drug and Alcohol Dependence, 112, 54-61.  PubMed ID 20580495

     A risk-taking "set" in a novel task among adolescents with serious conduct
     and substance problems.
     Crowley TJ, Raymond KM, Mikulich-Gilbertson SK, Thompson LL, Lejuez CW. (2006) 
     Journal of the American Academy of Child and Adolescent Psychiatry, 45, 175-183.   PubMed ID 16429088

     Changes in sensation seeking and risk-taking propensity predict increases in alcohol use
     among early adolescents
     MacPherson L, Magidson JF, Reynolds EK, Kahler CW, Lejuez CW (2010)
     Alcoholism: Clinical and Experimental Research, 34, 1400-1408.  PubMed ID 20491737     PubMed Central
 
     Positive and negative reinforcement underlying risk behavior in early adolescents
     MacPherson L, Reynolds EK, Daughters SB, Wang F, Cassidy J, Mayes LC, Lejuez CW (2010)
     Prevention Science, 11, 331-342.  PubMed ID 20309633     PubMed Central


BART Performance and Biological Outcomes

     Previous research has examined BART measures of risk taking in relation to functional polymorphism in the regulatory region (5-HTTLPR) of the human 5-HT transporter (5-HTT) gene. This research has found that carriers of the short 5-HTTLPR allele had significantly lower number of adjusted average number of pumps than those how where homozygotes of the long allele.     
     Imaging studies have reported that increased activation in mesolimbic and frontal brain regions during BART performance. In the fMRI adapted version of the test most similar to the traditional version of the BART, activation was observed in the midbrain, striatum, anterior cingulate cortex/medial frontal cortex, or dorsal lateral prefrontal cortex. When the task was modified to allow participants respnonse options to manipulate level of risk and the amount of virtual monetary reward for each balloon. In this condition, there was robust activation of dopamine rich mesolimbic structures and frontal target regions, including midbrain (ventral tegmental area, VTA), striatum (nucleus accumbens, globus pallidus nucleus caudate, and putamen), anterior insula, dorsal lateral prefrontal cortex, and anterior cingulate cortex/medial frontal cortex.
     Citations for Biological Studies:
     Genetic contributions of the serotonin transporter to social learning of fear
     and economic decision making
     Crişan LG, Pana S, Vulturar R, Heilman RM, Szekely R, Druğa B, Dragoş N, Miu AC (2009)
     Social Cognitive and Affective Neuroscience, 4, 399-408.  PubMed ID 19535614     PubMed Central

     Neural correlates of voluntary and involuntary risk taking in the human brain:
     an fMRI Study of the Balloon Analog Risk Task (BART).
     Rao H, Korczykowski M, Pluta J, Hoang A, Detre JA (2008)
     Neuroimage, 42, 902-910.  PubMed ID 18582578



Acknowledgements
 

Some material presented here was adapted by Carl Lejuez and edited by Charles Mathias from the forthcoming book chapter:
     Behavioral Measures of Risk-taking and their Relevance to Addictive Behaviors.
     Dahne, J., Richards, J. M., Ernst, M., MacPherson, L., & Lejuez, C. W. (in press)
     Wiley-Blackwell Handbook of Addiction Psychopharmacology (J. MacKillop & H. DeWit, Eds.).


For more information or to obtain this instrument, please contact Dr. Carl Lejuez at: clejuez@umd.edu