22. Four different forecasting methods were applied to a set of time-series data. The root mean square error (RMSE) was calculated for each method. Which of the following four RMSE values would be associated with the forecasting method that fit the data set best? (A) 6.21 (B) 2.33 (C) 2.00 (D) 1.96 23. During the past 8 years, the level of demand for a company's product has increased from 100 units per day to 137 units per day. What was the average annual compound growth rate over this period? (A) 37.00% (B) 4.63% (C) 4.01% (D) None of the answers is correct. 24. A firm currently has sales of 120 units per month. If the average growth rate of sales is 2% per month, approximately how many units will the firm be selling per month at the end of one year? (A) 152 (B) 149 (C) 144 (D) None of the answers is correct. 25. The forecast level of sales for the month of May was 240 units. Actual sales in May turned out to be 200 units. Use an exponential smoothing coefficient of 0.80 to forecast sales for June. (A) 230 (B) 220 (C) 210 (D) None of the answers is correct. 26. The level of sales in May, June, July, and August were 84,92,83, and 89 , respectively. What is the three-period moving average forecast of sales in September? (A) 88 (B) 87 (C) 85 (D) None of the answers is correct. 27. One advantage of estimating a function in which all variables have been transformed into their natural logarithms is that (A) the estimated coefficients measure slopes values. (B) problems with heteroskedasticity will be eliminated. (C) problems with multicollinearity will be eliminated. (D) the estimated slope coefficients are all elasticities