European Association for the Study of Gambling8th

European Conference

14-17 September 2010

From Habit to Addiction: A Study in Online Gambling Behavior

Bill Jolley, Ph.D.

Norwich University

Deborah Black, MD

University of Vermont

Research Aim and HypothesisAIM:

To enhance societal welfare by the pre-identification of at-risk gamblers using predictive behavioral markers.

PREMISE:

Behavioral addiction is a pathological form of habit.HYPOTHESIS:

Increased levels of the neurobiological trait of impulsivity will have a significant moderating effect on the predictability and level of behavioral addiction (Bechara et al., 2009).

RESEARCH VENUE:Online slot machine betting

Predictors of Gambling AddictionAddiction Risk

Ad

diction

Risk

Gam

blin

g H

abit

Neurobiological Marker of Addiction

Problem Problem GamblersGamblers

Intervention

Recreational Gamblers

At Risk Gamblers

Theoretical Background

HabitsHabits

are “sequential, repetitivesequential, repetitive, motor, or cognitive

behaviorsbehaviors

elicited by external or internal triggers that … can go to completion without constant conscious

oversight”

(Graybiel 2008).

ImpulsivityImpulsivity

is

“a predispositionpredisposition

toward rapid, unplanned reactions …

without regard to the negative

consequences”

and is the neurobiological trait most consistently linked to addiction (Moeller, Barratt et al. 2001).

Addictive gambling Addictive gambling is “a series of impulsive choicesimpulsive choices...” “

…whereby the problem gambler more frequently

selects the smaller, more probable immediate opportunity to obtain money . . . ”

resulting in more

disadvantageous outcomes. (Petry and Casarella 1999).

Hypothesized Model

HabitualBehavior

Sessions/days logged in Buy-In Amount

HabitualHabitualBehaviorBehavior

Sessions/days Sessions/days logged inlogged inBuyBuy--In AmountIn Amount

AddictiveOutcomesAmount BetWon/LostDuration

AddictiveAddictiveOutcomesOutcomesAmount BetWon/LostDuration

ImpulsivityIowa Gambling

Task

ImpulsivityImpulsivityIowa Gambling Iowa Gambling

TaskTask

The Iowa Gambling Task (Bechara et al. 1994)

High negative net scores indicate an impulsivity disorder where subjects make more choices that are disadvantageous.

Net score equals choices from advantageous decks minus choices from disadvantageous decks out of 100.

Research Design

Choice of 4 different slot games

Open 24/7 to university students only

February to May 2010

100 cash prize winners

Leader Board ranking determined prize amount

500 credits free, then 250 more credits for 2 hours of university / community service

To our knowledge, this is the first prospective, online

study of the effects of habit-inducing game characteristics and impulsivity on gambling behavior.

eCasinoLand Landing Page

eCasinoLand Game Page

The Sample

155 students registered

80% male; 83.9% had

gambled in past year

118 (76.1%) bet at least once

Stu

dent

s R

egis

tere

d

Calendar Days

Sequence of PlayDays Logged In

(n

= 118)

Number of Bets(n = 118)

IGT scores

Kolmogorov-Smimov

= .039

Median was -2.0

Cutoff for impulsive disorder < 10 (Bechara et al. 2004)

Impulsivity Did Not Effect the LEVELLEVEL

of Play

There were NO significant differences between high and low impulsivity groups on any

of the addictive gambling behavior measures!

IGT score and Playing Frequency

Global Path ModelResults support the hypothesis that gambling habit (defined by frequency and buy-in) is a predictor of addictive behavioral outcomes.

*Significant at p=0.05 level

0.587*0.443*

0.681*

FrequencyFrequencyAddictive

Behavioral Outcomes

R2 = 0.64

Addictive Behavioral Outcomes

R2 = 0.64Buy InBuy In

Interaction of Frequency &

Buy In

Interaction of Frequency &

Buy In

Moderation Effects of Impulsivity

Using the IGT median net score of -2 to test the moderation hypothesis produced two models with strong measurement reliability and strong construct validity, convergent validity and discriminant validity.

However:

There were no significant differences between impulsivity groups in the models’

hypothesized

relationships (path coefficients).

But, there was a significant reduction in models’

prediction quality (R2).

Low 

Impulsivity 

0.714*

0.05

0.283*

FrequencyFrequencyFrequencyAddictive

Behavioral Outcomes

R2 = 0.85

Addictive Behavioral Outcomes

R2 = 0.85Buy InBuy In

Interaction of Frequency &

Buy In

Interaction of Frequency &

Buy In

High 

Impulsivity

0.86*

‐0.236

0.062

FrequencyFrequencyAddictive Gambling Outcomes

R2 = 0.58

Addictive Gambling Outcomes

R2 = 0.58Buy InBuy In

Interaction of Frequency &

Buy In

Interaction of Frequency &

Buy In

Paradox or conundrum?

Impulsivity did not affect the level of addictive behavioral outcomes as hypothesized.

But it did affect the ability to predict them.

Therefore, we concluded that while impulsivity was not associated with more disadvantageous choices, it may have impaired the ability to make consistent choices.

Why the contrary results?

“Reward Deficiency”

hypothesis: Highly impulsive

individuals may require stronger stimuli (higher “entertainment value”) to develop a persistent habit.

Insufficient sample of players at the levels necessary to detect differences in behavior based on impulsivity.

Impulsivity is a risk factor in substance addiction, but this is the first study to examine the predictive value of the IGT in gambling behavior.

Impulsivity is not one-dimensional nor is it the only neurobiological factor related to addiction.