<p>@randombetch – You apparently don’t understand that the sample space of all kids who apply to college is not the same as the sample space of sophomores who attend Princeton and apply to WWS. So let’s use screwitlah’s argument that if the events were in the same sample space they would not be independent. The set of sophomores at Princeton who apply to WWU is contained in the set of applicants who were admitted to Princeton two years previously.</p>
<p>@Pea - you don’t know what a sample space is. Please take a stats class (or at least google the damn thing) and then come talk to me. You’re probably confusing “sample space” with “sample.”</p>
<p>You’re right, I confused random sample with sample space. You are mixing sample spaces. The sample space of all kids applying to colleges is not the same as the sample space of sophomores that attend Princeton. When you take two events from different random samples and multiply them together your result in meaningless.</p>
<p>Not necessarily. The admit rate for WWS isn’t going to chance in the future, so the chance of Bob getting into both WWS and Princeton, based purely on stochastic statistics, is 8% * 50% = 4%. Who says you have to mix the different samples?</p>
<p>Or you could mix the sample and still come up with a somewhat meaningful result.</p>
<p>Say the chance of Bob getting into Princeton is 8%.</p>
<p>Say the chance of Susan getting into WWS is 50%.</p>
<p>Say the chance of Alex getting into Harvard is 7%.</p>
<p>The chance of BOTH Bob getting into Princeton and Susan getting into WWS is 4%, which is lower than the chance of Alex getting into Harvard.</p>
<p>At Harvard any student can be a Government concentrator. The choice is made at the beginning of sophomore year. The Institute of Politics offers amazing leadership positions, grants and internships to all undergrads.</p>
<p>But I hope people will find this meaningful. For me it is!</p>
<p>Wow, I will definitely google the Institute of Politics from Harvard.</p>
<p>Thanks Fauve!</p>
<p>Also thanks Pea and randombetch!</p>
<p>This is a great discussion … But what about this:
Which is better for policy/IR/gov’t, WWS or Harvard Gov’t?</p>