7/1/25, 8:12
PM
ERM L2: z-test, effect size, ANOVAs, and Type 1 erros |
Exam Actual Questions and answers with complete
solutions verified latest update 2025/2026
(alpha)
rejecting a true null hypothesis / we reject the
Type I error null hypothesis when we shouldn't have
(false alarm)
(beta)
Type II error failing to reject a false null hypothesis /
alternative hypothesis is actually right, but we
didn't reject the null hypothesis (false
negative)
Correct decision: 1 - alpha we did not reject the null hypothesis
Correct decision: 1 - beta power!!!!!
probability of correctly rejecting the null hypothesis
when the alternative hyp is true
How do we Z-test!
determine/compute the
power?
1. Determine the Zcv for the given H0 (and the
assumed alpha & direction of a test)
Steps for determining 2. Determine the sample mean x̄ cv that belongs with
power: Zcv for the given H0
3. Convert the critical value Zh1 value for the given H1
4. The power is equal to chance: P(Z ≥ ZH1 | H1)
1/
6
, 7/1/25, 8:12
PM
Can we compute for one Yes, based on Warner 1, but its hard to compute
sample t-test? it when its not normally distributed, because
we need a computer application for this
what is critical value? The value beyond which you reject the null hypothesis.
It defines the cutoff for alpha
What determines that we alpha is small
get the largest probability power (1-beta) is large
of making correct
decisions?
- alpha
What are the four factors - sample size (large = narrows the distribution)
that influence power? - standard deviations (smaller = reduce noise)
- the 'true population mean' in the alternative
hypothesis (μh1)
The critical value line in the
distribution shifts to the left,
making the chances of
What happens if we rejecting the null
have a higher alpha hypothesis higher.
level? - power increases!
BUT, it comes with a cost!:
Type I error also increases
(incorrectly rejecting true
null hypothesis)
Describe what a normal
distribution with
μh0=0 and μh1 is like
2/
6
PM
ERM L2: z-test, effect size, ANOVAs, and Type 1 erros |
Exam Actual Questions and answers with complete
solutions verified latest update 2025/2026
(alpha)
rejecting a true null hypothesis / we reject the
Type I error null hypothesis when we shouldn't have
(false alarm)
(beta)
Type II error failing to reject a false null hypothesis /
alternative hypothesis is actually right, but we
didn't reject the null hypothesis (false
negative)
Correct decision: 1 - alpha we did not reject the null hypothesis
Correct decision: 1 - beta power!!!!!
probability of correctly rejecting the null hypothesis
when the alternative hyp is true
How do we Z-test!
determine/compute the
power?
1. Determine the Zcv for the given H0 (and the
assumed alpha & direction of a test)
Steps for determining 2. Determine the sample mean x̄ cv that belongs with
power: Zcv for the given H0
3. Convert the critical value Zh1 value for the given H1
4. The power is equal to chance: P(Z ≥ ZH1 | H1)
1/
6
, 7/1/25, 8:12
PM
Can we compute for one Yes, based on Warner 1, but its hard to compute
sample t-test? it when its not normally distributed, because
we need a computer application for this
what is critical value? The value beyond which you reject the null hypothesis.
It defines the cutoff for alpha
What determines that we alpha is small
get the largest probability power (1-beta) is large
of making correct
decisions?
- alpha
What are the four factors - sample size (large = narrows the distribution)
that influence power? - standard deviations (smaller = reduce noise)
- the 'true population mean' in the alternative
hypothesis (μh1)
The critical value line in the
distribution shifts to the left,
making the chances of
What happens if we rejecting the null
have a higher alpha hypothesis higher.
level? - power increases!
BUT, it comes with a cost!:
Type I error also increases
(incorrectly rejecting true
null hypothesis)
Describe what a normal
distribution with
μh0=0 and μh1 is like
2/
6
