# Estimating Monotonic Effects with brms

## Introduction

This vignette is about monotonic effects, a special way of handling discrete predictors that are on an ordinal or higher scale (Bürkner & Charpentier, in review). A predictor, which we want to model as monotonic (i.e., having a monotonically increasing or decreasing relationship with the response), must either be integer valued or an ordered factor. As opposed to a continuous predictor, predictor categories (or integers) are not assumend to be equidistant with respect to their effect on the response variable. Instead, the distance between adjacent predictor categories (or integers) is estimated from the data and may vary across categories. This is realized by parameterizing as follows: One parameter, $$b$$, takes care of the direction and size of the effect similar to an ordinary regression parameter, while an additional parameter vector, $$\zeta$$, estimates the normalized distances between consecutive predictor categories. For a single monotonic predictor, $$x$$, the linear predictor term of observation $$n$$ looks as follows:

$\eta_n = b D \sum_{i = 1}^{x_n} \zeta_i$

The parameter $$b$$ can take on any real value, while $$\zeta$$ is a simplex, which means that it satisfies $$\zeta_i \in [0,1]$$ and $$\sum_{i = 1}^D \zeta_i = 1$$ with $$D$$ being the number of elements of $$\zeta$$. Equivalently, $$D$$ is the number of categories (or highest integer in the data) minus 1, since we start counting categories from zero to simplify the notation.

## A Simple Monotonic Model

A main application of monotonic effects are ordinal predictors that can be modeled this way without falsely treating them either as continuous or as unordered categorical predictors. In Psychology, for instance, this kind of data is omnipresent in the form of Likert scale items, which are often treated as being continuous for convenience without ever testing this assumption. As an example, suppose we are interested in the relationship of yearly income (in $) and life satisfaction measured on an arbitrary scale from 0 to 100. Usually, people are not asked for the exact income. Instead, they are asked to rank themselves in one of certain classes, say: ‘below 20k’, ‘between 20k and 40k’, ‘between 40k and 100k’ and ‘above 100k’. We use some simulated data for illustration purposes. income_options <- c("below_20", "20_to_40", "40_to_100", "greater_100") income <- factor(sample(income_options, 100, TRUE), levels = income_options, ordered = TRUE) mean_ls <- c(30, 60, 70, 75) ls <- mean_ls[income] + rnorm(100, sd = 7) dat <- data.frame(income, ls) We now proceed with analyzing the data modeling income as a monotonic effect. fit1 <- brm(ls ~ mo(income), data = dat) The summary methods yield summary(fit1)  Family: gaussian Links: mu = identity; sigma = identity Formula: ls ~ mo(income) Data: dat (Number of observations: 100) Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1; total post-warmup samples = 4000 Population-Level Effects: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS Intercept 28.56 1.32 25.96 31.19 1.00 2794 2465 moincome 15.77 0.61 14.53 16.96 1.00 2702 2262 Simplex Parameters: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS moincome1 0.70 0.04 0.63 0.77 1.00 2808 2186 moincome1 0.17 0.04 0.08 0.26 1.00 3559 2308 moincome1 0.13 0.04 0.05 0.21 1.00 2991 1498 Family Specific Parameters: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS sigma 6.80 0.51 5.89 7.91 1.00 3300 2540 Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS and Tail_ESS are effective sample size measures, and Rhat is the potential scale reduction factor on split chains (at convergence, Rhat = 1). plot(fit1, pars = "simo") plot(conditional_effects(fit1)) The distributions of the simplex parameter of income, as shown in the plot method, demonstrate that the largest difference (about 70% of the difference between minimum and maximum category) is between the first two categories. Now, let’s compare of monotonic model with two common alternative models. (a) Assume income to be continuous: dat$income_num <- as.numeric(dat$income) fit2 <- brm(ls ~ income_num, data = dat) summary(fit2)  Family: gaussian Links: mu = identity; sigma = identity Formula: ls ~ income_num Data: dat (Number of observations: 100) Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1; total post-warmup samples = 4000 Population-Level Effects: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS Intercept 20.71 2.39 15.91 25.44 1.00 3920 2792 income_num 15.01 0.86 13.35 16.74 1.00 4067 3285 Family Specific Parameters: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS sigma 9.96 0.73 8.67 11.51 1.00 3879 3052 Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS and Tail_ESS are effective sample size measures, and Rhat is the potential scale reduction factor on split chains (at convergence, Rhat = 1). or (b) Assume income to be an unordered factor: contrasts(dat$income) <- contr.treatment(4)
fit3 <- brm(ls ~ income, data = dat)
summary(fit3)
 Family: gaussian
Links: mu = identity; sigma = identity
Formula: ls ~ income
Data: dat (Number of observations: 100)
Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
total post-warmup samples = 4000

Population-Level Effects:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept    28.39      1.32    25.81    31.02 1.00     2879     2512
income2      33.25      1.94    29.39    37.09 1.00     3239     2671
income3      41.09      1.98    37.11    45.09 1.00     3290     2981
income4      47.56      1.80    44.08    51.18 1.00     3043     2960

Family Specific Parameters:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     6.80      0.51     5.92     7.87 1.00     3614     2459

Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).

We can easily compare the fit of the three models using leave-one-out cross-validation.

loo(fit1, fit2, fit3)
Output of model 'fit1':

Computed from 4000 by 100 log-likelihood matrix

Estimate   SE
elpd_loo   -335.4  6.7
p_loo         4.8  0.7
looic       670.7 13.4
------
Monte Carlo SE of elpd_loo is 0.0.

All Pareto k estimates are good (k < 0.5).
See help('pareto-k-diagnostic') for details.

Output of model 'fit2':

Computed from 4000 by 100 log-likelihood matrix

Estimate   SE
elpd_loo   -372.6  6.6
p_loo         2.7  0.4
looic       745.2 13.2
------
Monte Carlo SE of elpd_loo is 0.0.

All Pareto k estimates are good (k < 0.5).
See help('pareto-k-diagnostic') for details.

Output of model 'fit3':

Computed from 4000 by 100 log-likelihood matrix

Estimate   SE
elpd_loo   -335.3  6.7
p_loo         4.8  0.7
looic       670.6 13.4
------
Monte Carlo SE of elpd_loo is 0.0.

All Pareto k estimates are good (k < 0.5).
See help('pareto-k-diagnostic') for details.

Model comparisons:
elpd_diff se_diff
fit3   0.0       0.0
fit1   0.0       0.1
fit2 -37.3       6.4  

The monotonic model fits better than the continuous model, which is not surprising given that the relationship between income and ls is non-linear. The monotonic and the unorderd factor model have almost identical fit in this example, but this may not be the case for other data sets.

## Setting Prior Distributions

In the previous monotonic model, we have implicitly assumed that all differences between adjacent categories were a-priori the same, or formulated correctly, had the same prior distribution. In the following, we want to show how to change this assumption. The canonical prior distribution of a simplex parameter is the Dirchlet distribution, a multivariate generalization of the beta distribution. It is non-zero for all valid simplexes (i.e., $$\zeta_i \in [0,1]$$ and $$\sum_{i = 1}^D \zeta_i = 1$$) and zero otherwise. The Dirichlet prior has a single parameter $$\alpha$$ of the same length as $$\zeta$$. The higher $$\alpha_i$$ the higher the a-priori probability of higher values of $$\zeta_i$$. Suppose that, before looking at the data, we expected that the same amount of additional money matters more for people who generally have less money. This translates into a higher a-priori values of $$\zeta_1$$ (difference between ‘below_20’ and ‘20_to_40’) and hence into higher values of $$\alpha_1$$. We choose $$\alpha_1 = 2$$ and $$\alpha_2 = \alpha_3 = 1$$, the latter being the default value of $$\alpha$$. To fit the model we write:

prior4 <- prior(dirichlet(c(2, 1, 1)), class = "simo", coef = "moincome1")
fit4 <- brm(ls ~ mo(income), data = dat,
prior = prior4, sample_prior = TRUE)

The 1 at the end of "moincome1" may appear strange when first working with monotonic effects. However, it is necessary as one monotonic term may be associated with multiple simplex parameters, if interactions of multiple monotonic variables are included in the model.

summary(fit4)
 Family: gaussian
Links: mu = identity; sigma = identity
Formula: ls ~ mo(income)
Data: dat (Number of observations: 100)
Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
total post-warmup samples = 4000

Population-Level Effects:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept    28.56      1.32    25.96    31.18 1.00     2810     2502
moincome     15.76      0.60    14.56    16.89 1.00     2552     2436

Simplex Parameters:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
moincome1     0.70      0.04     0.63     0.77 1.00     3374     2959
moincome1     0.16      0.04     0.08     0.25 1.00     3361     2246
moincome1     0.13      0.04     0.06     0.21 1.00     3139     1847

Family Specific Parameters:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     6.79      0.50     5.89     7.82 1.00     3322     2766

Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).

We have used sample_prior = TRUE to also obtain samples from the prior distribution of simo_moincome1 so that we can visualized it.

plot(fit4, pars = "prior_simo", N = 3) As is visible in the plots, simo_moincome1 was a-priori on average twice as high as simo_moincome1 and simo_moincome1 as a result of setting $$\alpha_1$$ to 2.

dat$age <- rnorm(100, mean = 40, sd = 10) We are not only interested in the main effect of age but also in the interaction of income and age. Interactions with monotonic variables can be specified in the usual way using the * operator: fit5 <- brm(ls ~ mo(income)*age, data = dat) summary(fit5)  Family: gaussian Links: mu = identity; sigma = identity Formula: ls ~ mo(income) * age Data: dat (Number of observations: 100) Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1; total post-warmup samples = 4000 Population-Level Effects: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS Intercept 29.71 3.67 22.85 37.10 1.00 1839 2366 age -0.03 0.08 -0.20 0.13 1.00 1960 2572 moincome 13.12 1.71 9.92 16.53 1.00 1256 2027 moincome:age 0.07 0.04 -0.01 0.15 1.00 1319 1923 Simplex Parameters: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS moincome1 0.77 0.08 0.64 0.93 1.00 1475 1679 moincome1 0.12 0.06 0.01 0.25 1.00 1694 1525 moincome1 0.11 0.06 0.01 0.22 1.00 2179 1823 moincome:age1 0.31 0.21 0.02 0.79 1.00 2423 2241 moincome:age1 0.37 0.22 0.02 0.83 1.00 2377 2372 moincome:age1 0.33 0.20 0.02 0.79 1.00 2662 2312 Family Specific Parameters: Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS sigma 6.71 0.49 5.85 7.76 1.00 3724 2911 Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS and Tail_ESS are effective sample size measures, and Rhat is the potential scale reduction factor on split chains (at convergence, Rhat = 1). conditional_effects(fit5, "income:age") ## Modelling Monotonic Group-Level Effects Suppose that the 100 people in our sample data were drawn from 10 different cities; 10 people per city. Thus, we add an identifier for city to the data and add some city-related variation to ls. dat$city <- rep(1:10, each = 10)
var_city <- rnorm(10, sd = 10)
dat$ls <- dat$ls + var_city[dat\$city]

With the following code, we fit a multilevel model assuming the intercept and the effect of income to vary by city:

fit6 <- brm(ls ~ mo(income)*age + (mo(income) | city), data = dat)
summary(fit6)
 Family: gaussian
Links: mu = identity; sigma = identity
Formula: ls ~ mo(income) * age + (mo(income) | city)
Data: dat (Number of observations: 100)
Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
total post-warmup samples = 4000

Group-Level Effects:
~city (Number of levels: 10)
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sd(Intercept)              10.97      3.32     6.16    18.99 1.00     1335     1940
sd(moincome)                1.13      0.79     0.06     2.98 1.00     1425     2028
cor(Intercept,moincome)    -0.48      0.47    -0.99     0.74 1.00     2977     2604

Population-Level Effects:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept       32.67      5.52    22.02    43.93 1.00     1608     1942
age             -0.01      0.10    -0.22     0.19 1.00     2341     2451
moincome        13.06      1.99     9.20    17.18 1.00     1824     2241
moincome:age     0.06      0.05    -0.04     0.16 1.00     1854     2405

Simplex Parameters:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
moincome1         0.77      0.08     0.63     0.93 1.00     2397     2153
moincome1         0.13      0.07     0.01     0.25 1.00     2650     1906
moincome1         0.11      0.06     0.01     0.23 1.00     2938     2053
moincome:age1     0.32      0.22     0.01     0.81 1.00     3229     2618
moincome:age1     0.36      0.22     0.02     0.82 1.00     3842     2545
moincome:age1     0.32      0.21     0.01     0.78 1.00     3715     2446

Family Specific Parameters:
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sigma     6.91      0.54     5.96     8.10 1.00     3892     2921

Samples were drawn using sampling(NUTS). For each parameter, Bulk_ESS
and Tail_ESS are effective sample size measures, and Rhat is the potential
scale reduction factor on split chains (at convergence, Rhat = 1).

reveals that the effect of income varies only little across cities. For the present data, this is not overly surprising given that, in the data simulations, we assumed income to have the same effect across cities.