Mixture models of delay discounting and smoking behavior

Abstract

The results inform the interpretation of steep discounting among smokers and suggest that treatment approaches could be tailored to the type of discounting behavior that smokers exhibit.

Discounting: A practical guide to multilevel analysis of indifference data

“…Nonlinear multilevel modeling is readily available for free in R , and analyzing the most common basic designs used in delay discounting research is well within the reach of every scientist. Although there are certain to be continued advancements in the analysis of indifference point data including the use of latent variables to identify important subgroups best fitted by different discounting functions (e.g., Hofmeyr et al, ) or bootstrapping of parameter estimates (Louie & Glimcher, ), currently available multilevel approaches offer too many benefits to delay their implementation.…”

A systematic review of temporal discounting in eating disorders and obesity: Behavioural and neuroimaging findings

“…Research of TD in relation to addictive behaviour has shown that increased rates of TD predict initiation and/or future use of cigarettes and cocaine in high-school students (Audrain-McGovern et al, 2009;Ayduk et al, 2000) . Moreover, individuals who smoke or are dependent on opioids, cocaine or methamphetamine (Hofmeyr et al, 2016;Mejia-Cruz et al, 2016;Robles et al, 2011) display increased rates of TD and higher rates of discounting is proposed to be an index of dependence severity (Amlung et al, 2016). Following a variety of treatments (e.g.…”

Modern statistical practices in the experimental analysis of behavior: An introduction to the special issue

“…Other statistical techniques are surfacing and bear watching. In particular, latent variable techniques including finite mixture models (Hofmeyr et al, 2016;Oberski, 2016) permit the modeling of continuous or categorical unobserved variables (i.e., intervening variables, MacCorquodale & Meehl, 1948). Unlike cluster analysis which groups observed data patterns based on their similarity, these latent variable approaches consider latent cluster membership to be probabilistic and thus provide stronger theoretical modeling of multimodal data like those commonly observed in behavioral preparations involving response rates and interresponse times (Li, Hautus, & Elliffe, 2018;Ray & McGill, 1964;Young, Vangsness, & McCoy, 2018).…”