H567/03 Section B Option 1 - Barkley-Levenson and Galván (2014) and reducing risk

H567/03 Section B Option 1 - Barkley-Levenson and Galván (2014) and reducing risk

Barkley-Levenson and Galván investigated how adolescents and adults represent the value of a risky choice. Their findings explain one aspect of brain development and help generate strategies that make rewarding opportunities safer.

The question about value

Previous research suggested heightened adolescent reward-system activity, particularly in the ventral striatum. However, monetary rewards might matter more to adolescents simply because they usually have less spending money. The researchers asked whether developmental differences remained when objective gamble value, income and observed acceptance behaviour were considered.

Subjective value is how attractive an outcome is to an individual. Expected value is the probability-weighted average of possible monetary outcomes. For outcomes xix_i with probabilities pip_i:

EV=∑ipixi.EV=\sum_i p_i x_i.

A 50% chance of gaining 20anda5020 and a 50% chance of losing 10 gives EV=0.5(20)+0.5(-10)=\5.Thisisalong−runaverage,notaguaranteed. This is a long-run average, not a guaranteed 5 payment. The same objective EV can have different subjective significance for different people.

The researchers predicted that increasing EV would influence adolescent gamble acceptance more strongly than adult acceptance; that ventral-striatum activation would be more sensitive to increasing EV in adolescents; and that the neural difference would remain even when adolescents were compared with adults showing similar acceptance behaviour. The topic is biological brain development, connected to cognitive valuation and decision making.

Sample, design and ethical preparation

The sample comprised 22 healthy right-handed adolescents aged 13–17, mean 15.6, with 11 female, and 19 healthy right-handed adults aged 25–30, mean 27.9, with 11 female. Participants were recruited through approved posters, internet advertisements and a database of previous participants at UCLA. They reported no psychiatric or neurological diagnosis or developmental delay, no body metal, and no psychoactive medication.

The age comparison was a laboratory quasi-experiment with independent groups: adulthood or adolescence was naturally occurring and not randomly assigned. Each participant also experienced gambles with different manipulated gain/loss values. Behavioural measures included acceptance and response time; fMRI measured blood-oxygenation changes associated with value processing. It is therefore incomplete to describe the whole investigation as only a measurement of gambling behaviour or only an independent-groups manipulation of every variable.

At an intake session, adults gave informed consent; under-18s gave assent and parents or guardians completed consent. Participants reported their monthly spending money, averaging 52.50foradolescentsand52.50 for adolescents and 467.11 for adults, and became familiar with the MRI environment using a mock scanner.

Participants received 20fortheintakesessionandweretolditwouldbetheirplayingmoneyaboutoneweeklater,withapossiblefurthergainorloss.Treatingthisasearnedmoneywasintendedtoreducethehouse−moneyeffect,inwhichpeopleriskmoneydifferentlywhentheyregarditasawindfall.Theywereactuallyassigneda20 for the intake session and were told it would be their playing money about one week later, with a possible further gain or loss. Treating this as earned money was intended to reduce the house-money effect, in which people risk money differently when they regard it as a windfall. They were actually assigned a 5–$10 payment corresponding to an accepted trial so they did not have to return money to the experimenters. This protected against financial loss but involved a discrepancy between the stated and actual payment procedure.

The gambling and scanning procedure

During fMRI, a spinner display offered a 50% probability of a stated gain and a 50% probability of a stated loss. Gain amounts ranged from +5to+5 to +20 and loss amounts from −5to−5 to −20. There were 144 mixed-gamble trials, with EVs from −7.50to+7.50 to +7.50. An additional 24 gain-only and 24 loss-only trials were interspersed, producing 192 trials across four runs. The gain-only/loss-only outcomes were known rather than risky; their EVs covered +6to+6 to +19 and −6to−6 to −19 respectively.

Participants decided whether they would accept each gamble for real money. They responded within 3,000 milliseconds using keys, and gamble outcomes were not revealed during scanning. They were told that one accepted trial would later be selected and played to determine payment. Gain/loss screen side and stimulus order were counterbalanced to reduce systematic presentation effects.

The task was adapted for younger participants: response options remained visible to reduce working-memory demands; white text avoided different colour cues for gains and losses; gain and loss ranges were equal; and participants received extensive training. These choices reduce the risk that a neural difference merely reflects misunderstanding or avoidable task difficulty.

Scanning used a 3-Tesla Siemens Trio MRI system. Functional signals are indirect measures associated with neural activity; they do not show that a region is the sole cause of a choice. Motion correction and standardised analysis were used. One adolescent was excluded for excessive movement, leaving 21 adolescents and 19 adults for scan analyses. The initial behavioural modelling reported three outliers excluded and 20 adolescents and 17 adults retained. These analysis-specific counts should be distinguished from the original 41 recruited; the brief outlier report does not fully describe the overlap of exclusions.

Continue this lesson.

You’re previewing NovaLearn without an account. Create a free account to read the rest.

Create free account