I've been looking at this proof for a while, and I think I was able to answer the first part of the question. It asks:

(i) For any events $X$ and $Y$ with $\textbf{P}(Y) > 0$, show that $\textbf{P}(X|Y) + P(X^{c}|Y) = 1$.

My proof is as follows:

  • $\textbf{P}(X|Y) + \textbf{P}(X^{c}|Y ) = \textbf{P}(A\cap B)/\textbf{P}(B) + \textbf{P}(A^{c}\cap B)/\textbf{P}(B)$
  • $\textbf{P}(A\cap B) + \textbf{P}(A^{c}\cap B) = \textbf{P}(B)$
  • $\textbf{P}(A\cap B)/\textbf{P}(B) + \textbf{P}(A^{c} \cap B)/\textbf{P}(B) = (\textbf{P}(A\cap B) + \textbf{P}(A^{c} \cap B))/\textbf{P}(B)$
  • So, $(\textbf{P}(A \cap B) + \textbf{P}(A^{c} \cap B))/\textbf{P}(B) = \textbf{P}(B)/\textbf{P}(B) = 1$

(ii) If $\textbf{P}(Y|X) > P(Y)$, show that $\textbf{P}(Y^{c}|X) < P(Y^{c})$ by using (i).

I have no idea how to prove this, though. Any help (on either part) would be greatly appreciated.

New contributor
TheFiveHundredYears is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
  • $\begingroup$ Have you tried proof by contradiction, say? $\endgroup$ – saulspatz Feb 14 at 0:55
  • $\begingroup$ @saulspatz No, I haven't. To be honest, I wouldn't really know a great place to start with that, either. $\endgroup$ – TheFiveHundredYears Feb 14 at 1:02
  • $\begingroup$ Why did you jump from using $X$ and $Y$ to using $A$ and $B$? Do not do that. $\endgroup$ – Graham Kemp Feb 14 at 1:07

For the second part, switch $X$ and $Y$. By assumption,

$$ \mathbb{P}(Y|X) > \mathbb{P}(Y) $$ which, by (i) $$ \Rightarrow 1 - \mathbb{P}(Y'|X) > \mathbb{P}(Y) \Rightarrow \mathbb{P}(Y'|X) < 1 - \mathbb{P}(Y) = \mathbb{P}(Y'). $$

  • $\begingroup$ Makes perfect sense, thank you! $\endgroup$ – TheFiveHundredYears Feb 14 at 1:21

Given $\def\P{\mathsf P} \P(Y)>0$ then your proof is:

$$\begin{align}\P(X\mid Y)+\P(X'\mid Y)&=\dfrac{\P(X\cap Y)}{\P(Y)}+\dfrac{\P(X'\cap Y)}{\P(Y)}&&\text{definition of conditioning}\\&=\dfrac{\P((X\cap Y)\cup(X'\cap Y))}{\P(Y)}&&\text{additivity over disjoint events}\\&=\dfrac{\P((X\cup X')\cap Y)}{\P(Y)}&&\text{distribution}\\&=\dfrac{\P(Y)}{\P(Y)}&&\text{union of complements}\\&=1 &&\text{since }\P(Y)>0\end{align}$$

Similarly you can show the following: $$\begin{align}\P(Y\mid X)+\P(Y'\mid X)&=1\\\P(Y)+\P(Y')&=1\end{align}$$

With a little rearrangement: $$\begin{align}\P(Y\mid X)&=1-\P(Y'\mid X)\\\P(Y)&=1-\P(Y')&\end{align}$$

Now, when $\P(Y\mid X)>\P(Y)$, then ...


Your Answer

TheFiveHundredYears is a new contributor. Be nice, and check out our Code of Conduct.

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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