revealPrefModel: Model Falsification with Random Limited Attention

In ramchoice: Revealed Preference and Attention Analysis in Random Limited Attention Models

Model Falsification with Random Limited Attention

Description

Given a collection of choice problems and corresponding choice probabilities, revealPrefModel determines if they are compatible with the Random Attention Model (RAM) of Cattaneo, Ma, Masatlioglu, and Suleymanov (2020) and/or the Attention Overload Model (AOM) of Cattaneo, Cheung, Ma, and Masatlioglu (2022).

See revealPref for revealed preference analysis with empirical choice data.

Usage

revealPrefModel(
  menu,
  prob,
  pref_list = NULL,
  RAM = TRUE,
  AOM = TRUE,
  limDataCorr = TRUE,
  attBinary = 1
)

Arguments

menu	Numeric matrix of 0s and 1s, the collection of choice problems.
prob	Numeric matrix, the collection of choice probabilities
pref_list	Numeric matrix, each row corresponds to one preference. For example, c(2, 3, 1) means 2 is preferred to 3 and to 1. When set to NULL, the default, c(1, 2, 3, ...), will be used.
RAM	Boolean, whether the restrictions implied by the RAM of Cattaneo et al. (2020) should be incorporated, that is, their monotonic attention assumption (default is TRUE).
AOM	Boolean, whether the restrictions implied by the AOM of Cattaneo et al. (2022) should be incorporated, that is, their attention overload assumption (default is TRUE).
limDataCorr	Boolean, whether assuming limited data (default is TRUE). When set to FALSE, will assume all choice problems are observed. This option only applies when RAM is set to TRUE.
attBinary	Numeric, between 1/2 and 1 (default is 1), whether additional restrictions (on the attention rule) should be imposed for binary choice problems (i.e., attentive at binaries).

Value

constraints	Matrices of constraints, generated by genMat. R: a matrix containing all constraints. ConstN: number of constraints for each preference.
inequalities	The moment inequalities. Positive numbers indicate that the RAM/AOM restrictions are rejected by the given choice probabilities. R: a vector containing all moment inequalities. ConstN: number of constraints for each preference.

Author(s)

Matias D. Cattaneo, Princeton University. [e-mail].

Paul Cheung, University of Maryland. [e-mail]

Xinwei Ma (maintainer), University of California San Diego. [e-mail]

Yusufcan Masatlioglu, University of Maryland. [e-mail]

Elchin Suleymanov, Purdue University. [e-mail]

References

M. D. Cattaneo, X. Ma, Y. Masatlioglu, and E. Suleymanov (2020). A Random Attention Model. Journal of Political Economy 128(7): 2796-2836. doi: 10.1086/706861

M. D. Cattaneo, P. Cheung, X. Ma, and Y. Masatlioglu (2022). Attention Overload. Working paper.

Examples

# Logit attention with parameter 2
# True preference: 1 2 3 4 5 6
menu <- prob <- matrix(c(1, 1, 1, 1, 1, 1,
                         0, 1, 1, 1, 1, 1,
                         1, 0, 1, 1, 1, 1,
                         1, 1, 0, 1, 1, 1,
                         1, 1, 1, 0, 1, 1,
                         1, 1, 1, 1, 0, 1,
                         1, 1, 1, 1, 1, 0), ncol=6, byrow=TRUE)
for (i in 1:nrow(prob)) prob[i, menu[i, ]==1] <- logitAtte(sum(menu[i, ]), 2)$choiceProb

# List of preferences to be tested
pref_list <- matrix(c(1, 2, 3, 4, 5, 6,
                      2, 3, 4, 5, 6, 1), ncol=6, byrow=TRUE)
# RAM only
result1 <- revealPrefModel(menu = menu, prob = prob, pref_list = pref_list, RAM = TRUE, AOM = FALSE)
summary(result1)

# AOM only
result2 <- revealPrefModel(menu = menu, prob = prob, pref_list = pref_list, RAM = FALSE, AOM = TRUE)
summary(result2)

# Both RAM and AOM
result3 <- revealPrefModel(menu = menu, prob = prob, pref_list = pref_list, RAM = TRUE, AOM = TRUE)
summary(result3)

ramchoice documentation built on May 24, 2022, 1:06 a.m.