Types of Error

Question 1

Why is there always a chance for error in NHST?

Answer 1

Because it relies on probability

Question 2

Type 1 error is what?

Answer 2

Rejecting the null when it’s actually true

Question 3

What is type 2 error?

Answer 3

You don’t reject the null when it’s actually true (false negative)

Question 4

In the boy who cried wolf, when everyone believed that there was a wolf present when there wasn’t, what kind of error is that?

Answer 4

Type 1

Question 5

In the boy who cried wolf, when people didn’t believe there was a wolf when there was one, what kind of error is that?

Answer 5

Type 2

Question 6

A false negative (type 2 error) occurs when you fail to find what?

Answer 6

Fail to find statistical significance for a real effect, caused by a small sample size

Question 7

In terms of population, what is a type 1 error (p

Answer 7

There is no difference in the population and null is true

Question 8

In terms of population, what is a type 2 error (p > .05)

Answer 8

There is a difference in population, null is false.

Question 9

Define effect size

Answer 9

An estimate of the size of an effect that is mostly independent of the sample.

Question 10

P0 value is very dependent on the ________ size

Answer 10

sample

Question 11

Cohen’s d and Pearson’s r are different types of _______ size.

Answer 11

effect

Question 12

Pearson’s r is the effect size of a ____________.

Answer 12

correlation

Question 13

What does is mean if 2 things are negatively correlated?

Answer 13

The higher you are on one scale the lower you are on another.

Question 14

What are the points of Cohen’s d?

Answer 14

The effect size of the difference between groups, determined by the size of the difference between groups and variability.

Question 15

T/F: Effect sizes and statistical significance are the same thing.

Answer 15

False, they are not the same. They are independent of one another

Question 16

What is power?

Answer 16

The probability that the null will be correctly rejected when false, and the ability to detect a statistically significant effect.

Question 17

What are some determinants of power?

Answer 17

Significance level
Smaller alpha level means less power
Effect size

Question 18

While lower alpha level decreases the likelihood of getting a false ________, they increase the likelihood of a false __________.

Answer 18

positive, negative

Question 19

Too much power increases likelihood of what kind of error?

Answer 19

Type 1 error

Question 20

What type of error contaminated literature with results that are not really there?

Answer 20

Type 1 error

Question 21

How do you determine necessary sample size before a study?

Answer 21

Conduct a power analysis. Also need to know effect size and planned p value you want to obtain.