You own a small store in a southeastern state where weather can be ex- treme, with frequent weather events producing high winds and thunder storms. This can knock power out to houses in your neighborhood.

You have looked at twelve years of data and found that during the months of February, March, and April you sell a lot of candles but then sales drop off. In fact, you have run a regression and discovered that sales are heaviest in March, drop off slightly for the following months, and reach a low in February.

Suppose the result of regression gives the equation

y = G(x) = 60 5x

where y represents sales of 100 boxes of candles x = month 2 so x is 1 for March, 2 for April, etc., reaching 11 for January and 12 for February. This means you sell 55 boxes in March and no boxes in February.

Calculating the correlation coefficient gives = .87.

1. [5 points] For the current year, would you expect to stock a lot of candles in January?

2. [5 points] Should you make decisions about numbers of boxes of can- dles to stock based on the month of the year?

Yes No

[10 points] Suppose at the end of April you have 30 boxes of candles in stock and you want to order enough to handle average sales in May. How many boxes of candles would you order