Allan’s Underground Systems installs septic systems for new houses constructed outside the city limits. To help forecast his demand, Mr. Allan has collected the data shown in Table on the number of country building permits issued per month, along with the corresponding number of bid requests he has received over a 15-month period.

Table 5.18 Month 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 No. Building permits 8 20 48 60 55 58 50 45 34 38 10 5 12 29 50 No. Bid requests 20 7 8 4 18 40 48 54 47 42 30 22 10 4 3

(a) Compute the simple correlation coefficient r between the number of building permits issued and the number of bid requests received in that month. Use all 15 periods of data.

(b) Use the first 12 months of data for building permits, and compute r between the number of building permits issued in a month and the number of bid requests received 2 months later (i.e., a 2-month lag).

(c) Repeat (b), but use a 3-month lag.

(d) Which type of regression model would be best to forecast bid requests: a same-month model, a 2-month lag model, or a 3-month lag model?

[Ans.

(a) 0.08

(b) 0.84

(c) 0.96

(d) A 3-month lag model is best. It permits Allan to explain 93 per cent of the variation in number of bid requests.]