Waterford Inc. is conducting cost-behaviour analysis on its selling and administrative expenses over a 15-month period. This was a period of significant growth for the company, and management would like to improve its ability to forecast these expenses for next year. The head office is in Vancouver and a cost analyst has selected number of customer orders per month as the cost driver. The following table and plot display the resulting monthly data.

A regression was run on these data, resulting in the following cost equation:
Y=$419+$2.32X
The R2 statistic for this regression is 0.71.

1. Calculate the squared error of the regression’s estimation of cost for the month of January 2011 only. (Hint: Refer to Exhibit 3-6. An error of a regression estimate is the vertical distance drawn between any cost observation and a point on the regression line at that level of activity.)

2. Explain the goodness of fit of this regression.

3. Is it correct to assume that this cost function is linear? Explain your answer fully.

4. The cost analyst ran another regression on the same 14 monthly observations of administrative cost data but with a different cost driver: direct material cost per month in the company’s production facility, located in Kelowna. Following are the regression results:

Y=$380+$2.75XR2=0.79

Would you recommend using the first regression for estimating administrative expenses (using customer orders per month as the cost driver), this one using direct material cost, or some other cost function? Give reasons to support your answer

Transcribed Image Text:

Activity Cost January 2011 500 February $1,500 550 1,800 March 600 2,000 April 675 2,200 May 750 2,325 June 875 2,400 July August September 1,150 October 1,000 2,500 1,050 2,625 2,875 1,250 3,125 November 1,300 3,250 December 1,400 3,500 January 2012 1,500 February 3,875 1,550 4,200 March 1,600 4,650 5,000 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 500 1,000 1,500 2,000 Activity Total Cost