2022-06-10 Psychologists speak another language
Today's question:
Psychologists speak another language
A group of psychologists is visiting the office, and Scarlett is in charge of showing them around.
The first stop will be in the Data Science department. They are very excited about showing them the results of their latest machine learning model.
Ten minutes into the presentation, it's painfully apparent that the crew is not fully grasping what is going on. Scarlett decides to summarize her ideas using a familiar language: statistics.
In statistics, the notion of statistical error is an integral part of hypothesis testing. There are two types of errors when testing the null hypothesis: type I and type II errors. Scarlett wants to explain their results regarding the latter.
Do you remember what the correct definition of a type II error is?
- [ ] A type II error occurs when the null hypothesis is true and is not rejected.
- [ ] A type II error occurs when the null hypothesis is true but is rejected.
- [ ] A type II error occurs when the null hypothesis is false but is not rejected.
- [ ] A type II error occurs when the null hypothesis is false and is rejected.
Let's review the answer:
It makes sense for those who are more used to machine learning terminology to compare type I and type II errors with false positives and false negatives.
Type I errors are the same as false positives. For example, if we mark a valid email as spam, we are in the presence of a false positive. Type I errors are the rejection of a true null hypothesis by mistake.
Type II errors are the same as false negatives. For example, if we let a spam message pass as a valid email, we are in the presence of a false negative. This is a type II error because we accept the conclusion of the email being good, even though it is incorrect. Type II errors are the acceptance of a false null hypothesis by mistake.
In other words, a type II error is when we incorrectly accept the null hypothesis even though the alternative hypothesis is true. Therefore, The third choice is the correct answer to this question.
Recommended reading
- "What Is a Null Hypothesis?" covers the basics you need to understand before going into hypothesis testing.
- Check out "Type I and type II errors" for the definition and examples of each type of error.
- "Understanding Null Hypothesis Testing" is an excellent article about hypothesis testing.