itchmodel is a computational model that can be used to understand the cognitive mechanisms underlying intertemporal choice performance: the defer/speedup effect (also known as the delay/speedup effect) and the date/delay effect.
The model builds on two tradeoff models of intertemporal choice: one that explains choice probabilities and full response time distributions in intertemporal choice Dai & Busemeyer, 2014 and another that can account for contextual effects, such as framing of time Scholten & Read, 2013.
Conceptually, the model explains intertemporal choice as a three-step process:
-
A decision maker transforms money into utility (value function), and calendar time into perceived/weighted time (time function), separately for the small-but-sooner and large-but-later options. These value and time functions are governed by power transformations.
-
Differences in utility are compared against differences in perceived/weighted time with different weights (w scales the difference between weighted delays and the difference between valued outcomes in a common currency), resulting in d, the overall advantage of the large-but-later option over the small-but-sooner option.
-
The mean rate of preference, d, is accumulated over time to a threshold that triggers an explicit preferential choice (i.e. diffusion or sequential-sampling process), providing predicted choice and response time.
The current version of the model can be parameterized in multiple ways to explain context effects, such as time framing:
-
Parameter mu varies between frames, corresponding to the hypothesis that time framing influences valuation.
-
Parameter kappa varies between frames, corresponding to the hypothesis that time framing influences time weighting/perception.
-
Parameters mu and kappa vary between frames, corresponding to the hypothesis that time framing influences both valuation and time weighting/perception.
-
Parameters mu and t0 vary between frames; as 1., but also explaining RT differences due to differences in stimuli.
-
Parameters kappa and t0 vary between frames; as 2., but also explaining RT differences due to differences in stimuli.
-
Parameters mu, kappa, and t0 vary between frames; as 3., but also explaining RT differences due to differences in stimuli.
The model can be installed from R as follows:
devtools::install_github("bramzandbelt/itchmodel")
library(itchmodel)
Usage will be explained in more detail soon. In the meantime, see the R Markdown notebooks inside the analysis directory for usage examples.
0.0.1 - June 2018
E-mail: bramzandbelt@gmail.com
-
Scholten, M. & Read, D. (2013) Time and outcome framing in intertemporal tradeoffs. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39, 1192–1212.
-
Dai, J. & Busemeyer, J.R. (2014) A probabilistic, dynamic, and attribute-wise model of intertemporal choice. Journal of Experimental Psychology: General, 143, 1489–1514.