What type of probability distribution is used for parts cost per unit?

• A. Normal distribution
• B. Uniform probability distribution
• C. Exponential distribution
• D. Binomial distribution

Answer: (B)

Costs per unit that can vary within a defined range are modeled with a Uniform distribution when there’s no reason to prefer any particular value inside that range. If you know the minimum and maximum possible cost and lack evidence suggesting a most likely value, treating every value in that interval as equally probable makes sense. In Monte Carlo simulations, you can sample by taking a random number uniformly distributed between 0 and 1 and mapping it to the [min, max] interval. This reflects the idea that the cost per unit could reasonably fall anywhere in that span with equal likelihood.

Why not other choices? A Normal distribution assumes data cluster around a central mean with probabilities falling off as you move away from the mean, which would imply a most likely cost and less likelihood for values far from it — not appropriate when you only know bounds. An Exponential distribution puts most weight near zero and has a long tail, implying costs are more likely to be small and rarely large, which isn’t consistent with a fixed interval [min, max]. A Binomial distribution models a count of successes in a fixed number of trials, not a continuous cost value.