Null Hypothesis and Significant Testing - Week Five

Question 1

What are the two possible conclusions from NHST

Answer 1

Observed Effect = Reflects Real Effect in Population
Observed Effect results from chance = doesn’t exist

Question 2

What are we looking for as a result when reviewing a hypothesis

Answer 2

Looking to reject H0
Can’t ever prove H1

Question 3

Define law of parsimony

Answer 3

Choose simplest explanation : chance

Question 4

How do we calculate and define the P Value?

Answer 4

P Value = P (Data / H0)
The probability of obtaining the observation or more extreme under H0

Question 5

Explain alpha level

Answer 5

Typically %%
If p-value < a = don’t reject H0 (statistically non-significant)
If p-value > a = reject H0 (statistically significant)

Question 6

Define unidirectional and bidirectional hypothesis?

Answer 6

Unidirectional = stating effect we will find in direction (big/small)
Bidirectional = Stating there is an effect but don’t know which direction

Question 7

Define one-tailed and two-tailed hypothesis?

Answer 7

One-tailed = is group 1 smaller than 2?
used if clear hypothesis before experiment and result in expected direction
p-value / 2 when one-tailed

Two-tailed= Is group 1 different to group 2?
used when conditions of one-tailed haven’t been met