1
$\begingroup$

i'm new at this, so i´m really sorry sorry if i do something wrong. The problem is this: A choice function satisfies condition α if whenever $x = C(A)$ and $x ∈ B ⊂ A$, it follows that $x = C(B)$ as well. Axiom 1: If $x ≠ C(A)$ then C(A) = C(A\{x}) And i need to show if condition α implies Axiom 1.

C(A) is choice function. Here is a definition of choice function (just in case you need it): "Let X be a finite set of alternatives. Let P(X) be the set of all nonempty subsets of X. We call the elements of P(X) as menus. A choice function is a map C : P(X) → X such that C(A) ∈ A, for all A ∈ P(X). It selects a single element from each menu. A choice correspondence is a map C : P(X) →P(X) such that C(A) ⊆ A, for all A ∈P(X). From each menu, at least one alternative is chosen"

And i found that Axiom 1 does not implies Condition α. Here is a counter example: We have X = {x, y, z}. Take A = {x, y, z} and y∈C({x, y, z}) y∈C({x, y}), z∈C({x, z}) and {y, z}∈C({ y, z}). Here we have a contradiction because C({x, y, z})≠C({ y, z}). Thank you.

Improve this question
$\endgroup$
4
  • $\begingroup$ Which condition is Condition 1? $\endgroup$ – Herr K. Nov 10 '17 at 0:45
  • $\begingroup$ im sorry. I meant condition α. Ty for letting me know that error. $\endgroup$ – neto333 Nov 10 '17 at 0:53
  • 1
    $\begingroup$ Something about Axiom 1 doesn't look right. Suppose $A=\{x,y,z\}$. Your Axiom 1 says $x=C(A)\;\Rightarrow\;C(A)=C(A\setminus\{x\})$, which means $x=C(\{y,z\})$?? $\endgroup$ – Herr K. Nov 10 '17 at 1:47
  • $\begingroup$ Another mistake. Sorry i should read my post more carefully next time, Axiom 1 should say $x≠C(A)$ then $C(A) = C(A\{x})$ .Sorry im having some trouble trying to writhe in code. $\endgroup$ – neto333 Nov 10 '17 at 1:59
2
$\begingroup$

Your "counterexample" also violates condition $\alpha$: $y=C(\{x,y,z\})$ but $y\ne C(\{y,z\})$. So it's not really a counterexample.

Improve this answer
$\endgroup$
5
  • 1
    $\begingroup$ @neto333: Your question says you want to "show if condition α implies Axiom 1", not the other way around. Also, your example does not satisfy Axiom 1 either. $\endgroup$ – Herr K. Nov 10 '17 at 2:39
  • $\begingroup$ Sorry, i think you are right. I completely forgot what i was trying to proof. We have that ${y, z}∈C({ y, z})$ so why this violates condition α $\endgroup$ – neto333 Nov 10 '17 at 2:49
  • $\begingroup$ @neto333: $\{y\}\ne\{y,z\}$. Note that the definitions use the "$=$" sign. $\endgroup$ – Herr K. Nov 10 '17 at 3:15
  • $\begingroup$ Ok, i think your answer is the answer. Just one more question before i mark this post as answered. I was not able to find any counter example (maybe im wrong) so can we say condition α implies Axiom 1? Or im i missing a counter example? $\endgroup$ – neto333 Nov 10 '17 at 3:18
  • $\begingroup$ @neto333: Yes, the implication can be proved mathematically. $\endgroup$ – Herr K. Nov 10 '17 at 3:23

Your Answer

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.