Cynthia J. Pietras

Relationships among laboratory and psychometric measures of impulsivity: Implications in substance abuse and dependence.

Objectives Problems with impulsive behavior are a key feature in substance use disorders. Many studies have examined this relationship, but methods used to measure impulsivity have varied greatly. The present study used a multimethod approach to determine relationships using both laboratory and psychometric measures of impulsivity commonly employed in the behavioral and health sciences. Methods Thirty-two adult male subjects participated for 5 days, completing psychometric instruments on day 1 and behavioral tests in counterbalanced order on days 2–5. Results Correlations between the behavioral and psychometric measures were uniformly low. Correlations within the psychometric instruments were consistently high. Principal components analysis revealed that the behavioral measures loaded into separate factors of response inhibition tasks and delay of reward tasks. Psychometric measures loaded into a single factor. Conclusions Results are discussed in terms of how studies of impulsivity might be interpreted based upon the tasks used, and how these interpretations may subsequently guide theory and measurement in substance abuse and dependence.

Response-cost punishment via token loss with pigeons

Two experiments were conducted to investigate punishment via response-contingent removal of conditioned token reinforcers (response cost) with pigeons. In Experiment 1, key pecking was maintained on a two-component multiple second-order schedule of token delivery, with light emitting diodes (LEDs) serving as token reinforcers. In both components, responding produced tokens according to a random-interval 20-s schedule and exchange periods according to a variable-ratio schedule. During exchange periods, each token was exchangeable for 2.5-s access to grain. In one component, responses were conjointly punished according to fixed-ratio schedules of token removal. Response rates in this punishment component decreased to low levels while response rates in the alternate (no-punishment) component were unaffected. Responding was eliminated when it produced neither tokens nor exchange periods (Extinction), but was maintained at moderate levels when it produced tokens in the signaled absence of food reinforcement, suggesting that tokens served as effective conditioned reinforcers. In Experiment 2, the effect of the response-cost punishment contingency was separated from changes in the density of food reinforcement. This was accomplished by yoking either the number of food deliveries per component (Yoked Food) or the temporal placement of all stimulus events (tokens, exchanges, food deliveries) (Yoked Complete), from the punishment to the no-punishment component. Response rates decreased in both components, but decreased more rapidly and were generally maintained at lower levels in the punishment component than in the yoked component. In showing that the response-cost contingency had a suppressive effect on responding in addition to that produced by reductions in reinforcement density, the present results suggest that response-cost punishment shares important features with other forms of punishment.

Human risky choice: delay sensitivity depends on reinforcer type.

The present study was designed to help bridge the methodological gap between human and nonhuman animal research in delay-based risky choice. In Part 1, 4 adult human subjects made repeated choices between variable-time and fixed-time schedules of 30-s video clips. Both alternatives had equal mean delays of 15 s, 30 s, or 60 s. Three of 4 subjects strongly preferred the variable-delay alternative across all conditions. In Part 2, these 3 subjects were then provided pairwise choices between 2 variable-time schedules with different delay distributions. Subjects generally preferred the variable-delay distributions with a higher probability of short-reinforcer delays, consistent with accounts based on nonlinear discounting of delayed reinforcement. There was only weak correspondence between experimental results and verbal reports. The overall pattern of results is inconsistent with prior risky choice research with human subjects but is consistent with prior results with nonhuman subjects, suggesting that procedural differences may be a critical factor determining risk-sensitivity across species.

Impulsiveness and other personality dimensions in substance use disorders and conduct disorder

ObjectivesThe psychometric measurement of personality and behavior traits may inform both basic research and clinical practice in substance abuse research. Relationships among novelty seeking, harm avoidance, reward dependence, impulsivity and venturesomeness, and conduct disorder were examined in a group of subjects with a past diagnosis of a substance use disorder and controls. MethodsPsychometric data from 68 subjects were analyzed using Analysis of Variance and logistic regression. ResultsStatistical analyses demonstrated that individuals meeting criteria for a past substance use disorder showed lower reward dependenceand greater impulsivity than controls, when controlling for the presence of conduct disorder. A substance use disorder x conduct disorder interactionwas found on the dimensions of harm avoidance and venturesomeness. ConclusionsLogistic regression revealed that a past substance use disorder could be predictedfrom the combination of two factors: low reward dependence and nonplanning impulsivity. Results are discussed in the context of previous data on personality dimensions and substance abuse.