Last answered:

07 Mar 2024

Posted on:

04 Mar 2024


Slight Miscommunication

Hi! I believe there is a slight miscommunication at point 5:24 in the video. Here, Ken explains why P(Offer | Masters = 0) is always equal to 0 by stating "[...] there is no datapoints in the dataset where someone's received a master's [degree] but gotten an offer". That is incorrect because we can see two instances where people in this dataset have both a master's degree and an offer.

I think the correct way of phrasing why this probability is always 0 is because the likelihood, P(Masters=0 | Offer=1) = 0. That value is multiplied to produce the final conditional probability making the entire product, P(Offer | Masters = 0) also 0. The reason why this likelihood is 0 is because "out of the datapoints within this dataset that ALREADY HAVE an offer (Offer = 1), there are no instances where a datapoint in that set DOES NOT have a master's degree (Masters = 0)". In other words, P(Masters=0 | Offer=1) = 0 therefore, P(Offer=1 | Masters=0) is also 0.

I hope that is helpful and please let me know if I have misinterpreted something relating to this issue. Thank you!

1 answers ( 0 marked as helpful)
Posted on:

07 Mar 2024



Thank you for your question! There seems to be a mix-up in the explanation. Actually, when we say P(Offer | Masters = 0) is 0, what we really mean is that in our data, everyone who got an offer already has a master's degree. So, there's no case where someone without a master's degree (Masters = 0) has gotten an offer. This is why we say this probability is 0. It's all about what's in our data: no offer without a master's degree. I hope this clears things up!


Submit an answer