¿Comparando coeficientes logísticos en modelos con diferentes variables dependientes?

Esta es una pregunta de seguimiento de la que hice hace un par de días . Siento que pone una inclinación diferente sobre el tema, así que enumeré una nueva pregunta.

La pregunta es: ¿puedo comparar la magnitud de los coeficientes entre modelos con diferentes variables dependientes? Por ejemplo, en una sola muestra, digamos que quiero saber si la economía es un predictor más fuerte de votos en la Cámara de Representantes o para el Presidente. En este caso, mis dos variables dependientes serían el voto en la Cámara (codificado 1 para demócrata y 0 para republicano) y voto para presidente (1 para demócrata y 0 para republicano) y mi variable independiente es la economía. Esperaría un resultado estadísticamente significativo en ambas oficinas, pero ¿cómo evalúo si tiene un efecto 'más grande' en una más que en la otra? Puede que este no sea un ejemplo particularmente interesante, pero tengo curiosidad por saber si hay una manera de comparar. Sé que uno no puede simplemente mirar el "tamaño" del coeficiente. Entonces, ¿Es posible comparar coeficientes en modelos con diferentes variables dependientes? Y, si es así, ¿cómo se puede hacer?

Si algo de esto no tiene sentido, avíseme. Todos los consejos y comentarios son apreciados.

La respuesta corta es "sí, puede", pero debe comparar los estimados de máxima verosimilitud (MLE) del "modelo grande" con todas las covariables en cualquier modelo ajustado a ambos.

Esta es una forma "cuasi formal" de obtener la teoría de probabilidad para responder a su pregunta

En el ejemplo, e Y 2 son el mismo tipo de variables (fracciones / porcentajes), por lo que son comparables. Asumiré que se ajusta el mismo modelo a ambos. Entonces tenemos dos modelos:$Y_{1}$$Y_{2}$

l o g ( p 1

$$M_{1}:Y_{1i}\sim Bin(n_{1i},p_{1i})$$ M2:Y2i∼Bin(n2i,p2i)log(p 2 $$log\left(\frac{p_{1i}}{1-p_{1i}}\right)=\alpha_{1}+\beta_{1}X_{i}$$ $$M_{2}:Y_{2i}\sim Bin(n_{2i},p_{2i})$$ $$log\left(\frac{p_{2i}}{1-p_{2i}}\right)=\alpha_{2}+\beta_{2}X_{i}$$

Entonces tiene la hipótesis que desea evaluar:

$$H_{0}:\beta_{1}>\beta_{2}$$

$\{Y_{1i},Y_{2i},X_{i}\}_{i=1}^{n}$

$$P=Pr(H_0|\{Y_{1i},Y_{2i},X_{i}\}_{i=1}^{n},I)$$

$H_0$

$$P=\int_{-\infty}^{\infty} \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} Pr(H_0,\alpha_{1},\alpha_{2},\beta_{1},\beta_{2}|\{Y_{1i},Y_{2i},X_{i}\}_{i=1}^{n},I) d\alpha_{1}d\alpha_{2}d\beta_{1}d\beta_{2}$$

La hipótesis simplemente restringe el rango de integración, por lo que tenemos:

$$P=\int_{-\infty}^{\infty} \int_{\beta_{2}}^{\infty} \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} Pr(\alpha_{1},\alpha_{2},\beta_{1},\beta_{2}|\{Y_{1i},Y_{2i},X_{i}\}_{i=1}^{n},I) d\alpha_{1}d\alpha_{2}d\beta_{1}d\beta_{2}$$

Debido a que la probabilidad es condicional en los datos, tendrá en cuenta los dos posteriores separados para cada modelo

$$Pr(\alpha_{1},\beta_{1}|\{Y_{1i},X_{i},Y_{2i}\}_{i=1}^{n},I)Pr(\alpha_{2},\beta_{2}|\{Y_{2i},X_{i},Y_{1i}\}_{i=1}^{n},I)$$

Now because there is no direct links between $Y_{1i}$ and $\alpha_{2},\beta_{2}$, only indirect links through $X_{i}$, which is known, it will drop out of the conditioning in the second posterior. same for $Y_{2i}$ in the first posterior.

From standard logistic regression theory, and assuming uniform prior probabilities, the posterior for the parameters is approximately bi-variate normal with mean equal to the MLEs, and variance equal to the information matrix, denoted by $V_{1}$ and $V_{2}$ - which do not depend on the parameters, only the MLEs. so you have straight-forward normal integrals with known variance matrix. $\alpha_{j}$ marginalises out with no contribution (as would any other "common variable") and we are left with the usual result (I can post the details of the derivation if you want, but its pretty "standard" stuff):

$$P=\Phi\left(\frac{\hat{\beta}_{2,MLE}-\hat{\beta}_{1,MLE}}{\sqrt{V_{1:\beta,\beta}+V_{2:\beta,\beta}}}\right)$$

Where $\Phi()$ is just the standard normal CDF. This is the usual comparison of normal means test. But note that this approach requires the use of the same set of regression variables in each. In the multivariate case with many predictors, if you have different regression variables, the integrals will become effectively equal to the above test, but from the MLEs of the two betas from the "big model" which includes all covariates from both models.

Why not? The models are estimating how much 1 unit of change in any model predictor will influence the probability of "1" for the outcome variable. I'll assume the models are the same-- that they have the same predictors in them. The most informative way to compare the relative magnitudes of any given predictor in the 2 models is to use the models to calculate (either deterministically or better by simulation) how much some meaningful increment of change (e.g., +/- 1 SD) in the predictor affects the probabilities of the respective outcome variables--& compare them! You'll want to determine confidence intervals for the two estimates as well as so you can satisfy yourself that the difference is "significant," practically & statistically.

I assume that by "my independent variable is the economy" you're using shorthand for some specific predictor.

At one level, I see nothing wrong with making a statement such as

X predicts Y1 with an odds ratio of _ and a 95% confidence interval of [ _ , _ ] while X predicts Y2 with an odds ratio of _ and a 95% confidence interval of [ _ , _ ].

@dmk38's recent suggestions look very helpful in this regard.

You might also want to standardize the coefficients to facilitate comparison.

At another level, beware of taking inferential statistics (standard errors, p-values, CIs) literally when your sample constitutes a nonrandom sample of the population of years to which you might want to generalize.

Let us say the interest lies in comparing two groups of people: those with $X_{1} = 1$ and those with $X_{1} = 0$.

The exponential of $\beta_{1}$, the corresponding coefficient, is interpreted as the ratio of the odds of success for those with $X_{1} = 1$ over the odds of success for those with $X_{1} = 0$, conditional on the other variables in the model.

So, if you have two models with different dependend variables then the interpretation of $\beta_{1}$ changes since it is not conditioned upon the same set of variables. As a consequence, the comparison is not direct...