Explain type one error in the null hypothesis?

Answer 1

See link. My answer goes a bit beyond your question. If the null hypothesis is right (which is our initial assumption is correct), then we should of course accept it. If our data has convinced us to reject the null hypotheis, that is in actuality correct, we have committed a type 1 error. This is a valid definition both in life situation and in statistical hypothesis testing. It is the a question of how much data do you need to convince you to change an opinion (a prior assumption). Type 1 errors are committed when people jump to conclusions based on little data. In situations where it is better to be extra cautious, such as a smoke detector goes off, that we need little data to change our initial opinion (building not on fire), because the harm in a type 1 error is far outweighed by the good if the alternative hypothesis is true (building on fire!). So fire alarm goes off and we leave the building. Type 2 errors is when you need a lot of convincing before you change your opinion, or you fail to reject the null hypothesis (building not on fire) when it is false (alternative hypothesis, building on fire). In the fire, you don't leave until you see more data, ie. flames. Big mistake if there's no time left to leave. The Swine Flu epidemic offers an excellent example where type 1 errors (false positives) are much preferred over type 2 errors (false negatives). Are Type 2 error consequences worse? No. Suppose you are on trial. You would hope that the jurors would minimize type 1 error as they are suppose to do, at the risk of committing a type 2 error (free a guilty man).