What error type 1 or 2 should professionals avoid?

Answer 1

It depends on the circumstances. There are certain professions/ situations in which these errors play a role. First, I've included a related link. Alpha or Type I are false positives and Type II are false negatives. If the status quo is "good product" and the alternative is "bad product", then the Type 1 and 2 errors are rejecting good product, and accepting bad product, respectively. If we have very high standards, we increase the chance of throwing out good product. This is an area where statistics (hypothesis testing) and human behavior and attitudes overlap. Actual decision making is not as simple as the binary options (null and alternative hypothesis) in statistics. As a consumer, we generally should avoid type II errors, which come by not rejecting the null hypothesis (status quo) in light of new evidence. While avoiding type II errors, type 1 errors are increased. I'll give a simple example. I'm at a restaurant and a big juicy steak is put in front of me, but something smells a bit funny, so I reject the null hypothesis (good steak) in favor of the alternative (spoiled meat). I may have wasted my money on the steak, but at least I didn't take a chance of getting sick. I did not want to make a type II error (accepting bad steak) so I increased the chance of a type I error (rejecting a good steak). Well, life's not about binary decisions, perhaps I'll feed the steak to my dog. If a fire alarm goes off, people generally don't hang around to see smoke and flames. They quickly reject the null hypothesis in favor of the alternative (fire!!!). Those running out of the building increase the chance of type 1 errors (conclusion of fire when in fact, the alarm is faulty) to minimize the type 2 errors. In the medical profession, generally efforts are made to minimize type 2 errors, reacting on little information to change from status quo. This increases type 1 error. The legal system is a bit more complicated. The court system is designed so the chance of type 1 errors (a guilty finding of innocent people) must be less than type 2 error (a finding of not guilty of criminals). How about hiring? Companies do both errors. Let us say that the status quo (null hypothesis) is the person is not qualified. With low standards, they are likely to commit Type 1 errors (accept person as qualified, given they are not) and with high standards, they are likely to commit Type 2 error (accept the person as unqualified, given they are actually qualified).