Statistical Power Flashcards

(8 cards)

1
Q

What is statistical power?

A

The probability that a test correctly detects a real effect — i.e., it rejects a false null hypothesis.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What kind of error does statistical power help avoid?

A

Type II error (failing to detect a real effect).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

How does effect size (ES) influence power?

A

Larger effect sizes are easier to detect, so power increases with effect size.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

How does sample size (n) influence statistical power?

A

Larger sample sizes increase power by improving precision.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How does background variation (σ or σ²) affect power?

A

More variation lowers power because it makes real effects harder to detect.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is a common value for significance level (α), and what happens if you increase it?

A

Commonly α = 0.05; increasing it (e.g. to 0.10) increases power but also raises the chance of a Type I error.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

How does the type of statistical test used affect power?

A

Some tests are more sensitive or better suited to the data, which can increase or decrease power.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Why is adjusting α important in studies with low replication, like environmental impact assessments?

A

Because power is low in such studies, increasing α (e.g. to 0.10) helps detect real effects more reliably.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly