APPLICATION OF TYPE-I AND TYPE-II ERRORS IN DAIRY FARM-MANAGEMENT DECISION-MAKING

The economic costs of type I and type II errors associated with decisions regarding the use of sodium bicarbonate and bST on dairy farms were calculated. Data from published experiments were used to estimate the mean and SD in production responses and associated changes in DM intake. The frequency distribution of the production response and the expected monetary value of type I error (using a product when it is unprofitable) and type II error (not using a product when it is profitable) were calculated. Supplemental sodium bicarbonate ($.05/cow per d) and bST ($.40/cow per d) were used as sample decision problems. Milk was valued at $.28/kg, and feed was priced at $.15/kg. Based on 12 reports, cows supplemented with .75% sodium bicarbonate in ration DM produced an additional 1.43 kg/d of 3.5% FCM (SD = 1.13) and consumed .36 kg/d more feed. Type I error costs were $.02/d, whereas type II error costs were $.32/d. From 16 reports, cows supplemented daily with 10 to 15 mg/d of bST produced an additional 3.9 kg/d of 3.5% FCM (SD = 2.1) and consumed .6 kg/d more feed. Type I error costs were $.03/d, whereas type II error costs were $.66/d. Because type II error costs were greater than type I error costs, the economic decision favored the use of sodium bicarbonate or bST or both. Type II error cost was greater than type I for both products when DM cost was varied from $.064 to .239/kg and milk price was varied from $.145 to $.429/kg.