Two employees can each work three of four contracted days and still have completely different attendance rates when the length of those days is considered.

Consider a contract for Monday and Tuesday from 9:00 to 16:00, Wednesday from 9:00 to noon, and Saturday from 10:00 to 14:00. After breaks, that is about nineteen hours across four working days.

Pattern one: with approval, the employee works Monday through Wednesday and takes Saturday off. Counting days gives 75 percent, but counting hours gives about 84 percent.

Pattern two: because the busy period changed, the workplace asks the employee to work Friday instead of Monday without formally updating the contract. A narrow comparison against the original contracted slots can produce 75 percent by days and roughly 66 percent by hours, even though the employee followed the company’s request.

If the contract had been updated to match either approved arrangement, the appropriate result could be 100 percent. The examples show why a system cannot blindly treat a contract as a perfect picture of reality.

Would a part-time worker accept losing paid-leave eligibility after working exactly as the workplace requested? A defensible system has to see both the contract and the actual work.

WRITTEN BYYoshio Taki

A systems engineer who loves IBM i / AS/400

CHAPTER 03 · PART 03 · EPISODE 2 / 10 STORIES · 35 STORIES TOTAL

Chapter 3 index · All 35 stories