---
title: "Modeling binary choice data"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Modeling binary choice data}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

```{r setup}
library(tempodisco)
```

## Classic methods

For binary choice data not explicitly designed to [titrate out indifference points](https://kinleyid.github.io/tempodisco/articles/adjusting-amounts.html) (as in an adjusting amount procedure), there are a few widely-used traditional scoring methods to quantify discounting.

### Kirby scoring

One scoring method is the one designed for the Monetary Choice Questionnaire [(Kirby, 1999)](https://doi.org/10.1037//0096-3445.128.1.78):

```{r}
data("td_bc_single_ptpt")
mod <- kirby_score(td_bc_single_ptpt)
print(mod)
```

Although this method computes $k$ values according to the hyperbolic discount function, in principle it's possible to use other single-parameter discount functions (though this is not an established practice and should be considered an experimental feature of `tempodisco`):

```{r}
mod_exp <- kirby_score(td_bc_single_ptpt, discount_function = 'exponential')
print(mod_exp)
mod_pow <- kirby_score(td_bc_single_ptpt, discount_function = 'power')
print(mod_pow)
mod_ari <- kirby_score(td_bc_single_ptpt, discount_function = 'arithmetic')
print(mod_ari)
```

### Wileyto scoring

It is also possible to use the logistic regression method of [Wileyto et al. (2004)](https://doi.org/10.3758/BF03195548), where we can solve for the $k$ value of the hyperbolic discount function in terms of the regression coefficients:

```{r}
mod <- wileyto_score(td_bc_single_ptpt)
print(mod)
```

## Newer methods

### Linear models

The [Wileyto et al. (2004)](https://doi.org/10.3758/BF03195548) approach turns out to be possible for other discount functions as well ([Kinley, Oluwasola & Becker, 2025](https://doi.org/10.1016/j.jmp.2025.102902).):

```{r child="../man/fragments/linear-models.Rmd"}
```

We can test all of these and select the best according to the Bayesian Information Criterion as follows:

```{r}
mod <- td_bclm(td_bc_single_ptpt, model = 'all')
print(mod)
```

### Nonlinear models

To explore a wider range of discount functions, we can fit a nonlinear model by calling `td_bcnm`. The full list of built-in discount functions is as follows:

```{r child="../man/fragments/predefined-discount-functions.Rmd"}
```

```{r}
mod <- td_bcnm(td_bc_single_ptpt, discount_function = 'all')
print(mod)
```

#### Choice rules

Several additional arguments can be used to customize the model. For example, we can use different choice rules---the "logistic" choice rule is the default, but the "probit" and "power" choice rules are also available (see [this tutorial](https://kinleyid.github.io/tempodisco/articles/choice-rules.html) for more details):

```{r}
# Probit choice rule:
mod <- td_bcnm(td_bc_single_ptpt, discount_function = 'exponential', choice_rule = 'probit')
# Power choice rule:
mod <- td_bcnm(td_bc_single_ptpt, discount_function = 'exponential', choice_rule = 'power')
```

#### Error rates

It is also possible to fit an error rate $\epsilon$ that describes the probability of the participant making a response error (see [Vincent, 2015](https://doi.org/10.3758/s13428-015-0672-2)). I.e.: $$P(\text{imm}) = \epsilon + (1 - 2\epsilon) g^{-1}[\eta]$$ where $P(\text{imm})$ is the probability of choosing the immediate reward, $g$ is the link function, and $\eta$ is the linear predictor.

```{r}
data("td_bc_study")
# Select the second participant
second_ptpt_id <- unique(td_bc_study$id)[2]
df <- subset(td_bc_study, id == second_ptpt_id)
mod <- td_bcnm(df, discount_function = 'exponential', fit_err_rate = T)
plot(mod, type = 'endpoints', verbose = F)
lines(c(0, 1), c(0, 0), lty = 2)
lines(c(0, 1), c(1, 1), lty = 2)
cat(sprintf("epsilon = %.2f\n", coef(mod)['eps']))
```

We can see that the probability of choosing the immediate reward doesn't approach 0 or 1 but instead approaches a value of $\epsilon \approx 0.11$.

#### Fixed endpoints

Alternatively, we might expect that participants should never choose an immediate reward worth 0 and should never choose a delayed reward worth the same face amount as an immediate reward ([Kinley, Oluwasola & Becker, 2025](https://doi.org/10.1016/j.jmp.2025.102902); see [here](https://kinleyid.github.io/tempodisco/articles/choice-rules.html#fixed-endpoint-choice-rules) for more details). We can control this by setting `fixed_ends = T`, which "fixes" the endpoints of the psychometric curve, where `val_imm = 0` and `val_imm = val_del`, at 0 and 1, respectively:

```{r}
mod <- td_bcnm(df, discount_function = 'exponential', fixed_ends = T)
plot(mod, type = 'endpoints', verbose = F, del = 50, val_del = 200)
```
