Deriving quantile of random variable which is a function of Normal distribution

31 Views Asked by At

Let say $X \sim N \left(0, 1\right)$

and, $Y = \Phi\left( \frac{a-bX}{c} \right), a,b,c, >0$, which $\Phi$ is the cdf of Standant Normal distribution.

I wish to calculate the ${\alpha}^{th}$ quantile of $Y$.

Can it be said that the ${\alpha}^{th}$ quantile of $Y$ is $\Phi\left( \frac{a-b X_{{\alpha}^{th}}}{c} \right)$, where $X_{{\alpha}^{th}}$ is the ${\alpha}^{th}$ quantile of the random variable $X$?

Any pointer will be very helpful

1

There are 1 best solutions below

0
On BEST ANSWER

Let $y_{\alpha}$ be the $\alpha-$quantile of $Y$. By the definition of the quantile,

$$\mathbb P\left[Y > y_\alpha\right] = \alpha\quad \implies \quad\mathbb P\left[X < \frac{a - c\Phi^{-1}\left(y_\alpha\right)}{b}\right] = \alpha \quad \implies \quad \mathbb P\left[X > \frac{a - c\Phi^{-1}\left(y_\alpha\right)}{b}\right] = 1-\alpha$$ So again by the definition of the quantile, $$\frac{a - c\Phi^{-1}\left(y_\alpha\right)}{b} = x_{1-\alpha}$$ finally $$y_\alpha = \Phi\left(\frac{a - b x_{1-\alpha}}{c}\right)$$