Questions $5-10$ are based on the Excel file "Minitest$4$DataQ$5$thru$10$Bankloans.x$|$sx$".$

In auditing you can ask two managers at an audit company about the chance that the

warranty reserve is adequately funded in relation to the increase in the company's sales and

get two very different forecasts regarding the probability that the funding level of the

warranty reserve is sufficient. This audit situation analogizes nicely to a business banking

situation: Will a business bank customer pay their loan payment this month or not?

Ying Li and Liam Smith differ on which model to utilize when predicting if $30$ of the bank's

business loans will not pay their January $2025$ loan payment. Ying and Liam differ on which

predicting model is better. Liam swears by the predicting model used by the East office

location of the bank. Ying thinks more accurate loan payment predictions can be made by

using the predicting model of the bank's West location office. They agree that it would help to

look at actual data to see who is right. So$,$ they look at the month of January.

Liam says the bank did not receive payments from $12$ of these $30$ business loan customers

during the month and that the two predicting models are equally accurate because their

average forecast is a $50\%$ probability of loan payment received across the $30$ customers. That

means each model predicted loan payments to not be received from $15$ of these business

customers, as $50\%$ of $30$ is $15.$ So$,$ each model under predicted who was not going to pay in

January by $3$ business customers, $(15$ minus $12).$ He also says that being off by $3$ business

customers is simply being more cautious as the bank reserves more funds for its loan loss

reserve, and then they should be be happy when more business loan customers pay. Ying

counters that it's more important, in her book, to not underestimate loan payments received,

as that receipt of payments impacts the cash flow of the bank and its mission to loan to

customers within the community to help local businesses. She finally concludes, "Overall,

they're equally accurate, and we should leave it at that."

Q$5.$ Which modeling method $($East or West$)$ is more accurate in predicting which loan

payments were not received in January?

$($Hint: I recommend that you download the file and familiarize yourself with it$.$ You will be

doing some computations and such with the data! You would also likely benefit if you

converted the "Loan payment received in January" column of "Received" and "Not Received"

into a new column of $2\text{'}$s and $1\text{'}$s$.)$