Type I and II Errors
1. A Type I Error in statistics, in layman’s term, equates to having rejected a hypothesis of interest while the assumption is actually true. On the other hand, a Type II error is committed when the hypothesis of interest is actually false but was failed to get rejected based on the underlying analysis (Intuitor, 1996).
On a personal experience in working with an IT distributor company, these errors were also observed in many different aspects of business transactions. One example when a Type I Error occurred was when the sales department actually predicted that one retailer was going to order for five desktop computers after three business days but failed to request from the inventory department to reserve stocks and instead released the remaining items to outright purchases from the store. In the end, the retailer actually ordered five desktops three days later but no more stocks were available.
Meanwhile, Type II Errors were also committed in the same company where I have previously worked for. This happened when one account number for a freight company was lost in the records. The inventory department assumed that the numbers indicated in the purchase order was the billable account number. One team suggested that it didn’t look any similar to other account numbers but the Release Department staff proceeded using it. In the end, the numbers were really not freight account numbers and the freight company charged the business instead.
2. In practice, Type II Errors seem to have graver effects to the company’s profitability than committing Type I Errors. Usually, non-measurable opportunities are lost when a Type I Error is committed but this does not readily affect the company’s money assets. On the other hand, type II errors are somehow preventable and could have been so if only clearer judgments and analyses were first executed over decision making. Moreover, these errors involve the loss of money and funds. In a bigger picture, Type I Errors are more tolerated than Type II Errors. These two error types almost always happen at the same rate but it seems that Type I Errors are only tolerated at least 25% of the time. This assumption probability may be based on the fact that the business imposes three times as large penalty to the staff when committing Type II Errors.
Intuitor, (1996). Type I and Type II Errors – Making Mistakes in the Justice System. Retrieved May 7, 2009, from Intuitor.com Web site: http://www.intuitor.com/statistics/T1T2Errors.html