2
$\begingroup$

I have a dependent variable named "FGR" which represents a ratio and takes values between 0 and +Inf. The histogram plot below illustrates the distribution of FGR, excluding +Infinite values

enter image description here

To categorize the FGR variable, I created a new ordinal variable called "FGR_cat" with three categories:

fit1 (<0.30) fit2 (>= 0.30 & < 0.80) fit3 (>= 0.80) Next, I conducted a generalized additive model using the ocat family (ordinal logistic regression). The model formula is as follows:


gam_model <- mgcv::gam(FGR_cat ~ s(Hormone) + Age + BMI + as.factor(PCOS) + s(Patient, bs = "re")
                       , data = my_data,
    ,method = "REML", family = ocat(theta = 1))

The model results for the effect of hormone are as follows:

Family: Ordered Categorical(-1,0)

Link function: identity

edf Ref.df Chi.sq p-value

s(Hormone) 1.0000083 1 4.12 0.0424

Deviance explained = 2.06%

-REML = 419.06 Scale est. = 1 n = 351 AIC value: 829.7235

The stacked area chart below shows the predicted probabilities of fit1, fit2, and fit3 along the hormone axis:

enter image description here

However, when I set the threshold for the linear predictor to 5 (theta = 5) using the following code:

,method = "REML", family = ocat(theta = 5))

The updated results are as follows:

Family: Ordered Categorical(-1,4) Link function: identity

edf Ref.df Chi.sq p-value

s(Hormone) 1.0 1 6.911 0.00857 **

Deviance explained = 71.4%

-REML = 359.13 Scale est. = 1 n = 351 AIC value = 646.0894

enter image description here

As the value of theta increases, the significance of the effect of Hormone becomes more pronounced, and the AIC consistently decreases. Eventually when theta reaches 10 or higher, no predicted probability for fit3 is observed at any level of Hormone.

By adjusting the value of the theta threshold, it is possible to obtain a wide range of predicted probabilities. However, determining the appropriate threshold values for the linear predictor that can be confidently reported in scientific studies is crucial. What is the most accurate approach for defining these threshold values? Alternatively, if you have any alternative methods for modeling the FGR ratio, please provide an explanation. Thank you

EDIT after the comment of @Paul: In the explanation of theta it is written: "cut point parameter vector (dimension R-2). If supplied and all positive, then taken to be the cut point increments (first cut point is fixed at -1). If any are negative then absolute values are taken as starting values for cutpoint increments"

I could not understand the meaning of last sentence. When i choose theta as positive, such as 3, it makes an increment in the value of first cutpoint -1 by 3 and the result is -1 + 3 = 2 as follows;

Family: Ordered Categorical(-1,2)

When i choose the theta as a negative value, it defines the cutoff values as 1.62 as follows;

Family: Ordered Categorical(-1,1.62)

How did it choose the threshold of 1.62?

$\endgroup$
5
  • 1
    $\begingroup$ According to mgcv documentation, theta will be estimated by gam if you don't specify it, so why not do that? rdrr.io/cran/mgcv/man/ocat.html $\endgroup$
    – Paul
    Commented Jun 7, 2023 at 12:01
  • $\begingroup$ @Paul Thank you very much. When I wrote "family = ocat()" it gives the following error: Error in ocat() : Must supply theta or R to ocat Please check the edit $\endgroup$
    – insan
    Commented Jun 7, 2023 at 12:23
  • 1
    $\begingroup$ So set R to be the number of categories you have in the response and leave theta unset and gam() will estimate the theta for you $\endgroup$ Commented Jun 7, 2023 at 13:35
  • $\begingroup$ @GavinSimpson Omg. That's absolutely true :) One more question. How should I describe the methodology of this technique (the model and the defined thresholds) in a scientific paper? Why is -1 chosen as the starting threshold? Is there a specific rationale behind it? Since my primary discipline isn't statistics, it's challengng to comprehend the statistics paper that introduced the "ocat" technique. For instance, if all covariates in the gam model have an EDF of 1, would the predictions based on different hormone levels be similar to the estimates obtained from proportional logistic regression? $\endgroup$
    – insan
    Commented Jun 7, 2023 at 14:36
  • $\begingroup$ Does this help with the interpretation: stats.stackexchange.com/a/234833/1390 You want to be looking for an ordered categorical response model where the linear predictor is a latent variable on the real number line with thresholds indicating the boundaries on the latent variable between classes. $\endgroup$ Commented Jun 9, 2023 at 7:07

1 Answer 1

2
$\begingroup$

Unless you have very strong prior knowledge, it is better to estimate the threshold values, $\boldsymbol{\theta}$. In that case you must specify the number of categories, $R$, via argument R to the ocat() family, and leave the theta argument set at its default NULL value.

By way of example, here is the model I use for testing in {gratia}:

library("gratia")
library("mgcv")

# Ordered categorical model ocat()
n_categories <- 4                                            # <--- How many categories

# simulate data
su_eg1_ocat <- data_sim("eg1", n = 200, dist = "ordered categorical",
  n_cat = n_categories, seed = 42)

# fit model to these data
m_ocat <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),
  family = ocat(R = n_categories),           # <--- specify R, number of categories here
  data = su_eg1_ocat, method = "REML")
```
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.