Null Hypothesis Testing Flashcards
Name the 2 hypotheses that Null Hypothesis Testing involves.
Null Hypothesis
Alternative Hypothesis
H0 = _____________
null hypothesis
H1= ________
alternative hypothesis
What is the null hypothesis?
Statement about the independence or equality in populations. Essentially saying nothing is happening, no effect.
What is the alternative hypothesis?
Statement about the presence of important differences or associations in populations. Essentially saying there is some kind of effect.
What hypothesis do we always favour?
The null hypothesis.
What hypothesis is this?: "there is no difference between results of men and women in the class test"
Null Hypothesis
What hypothesis is this?: "There is a difference between results of men and women in class test"
Alternative Hypothesis
We calculate the _____ that the difference between men and women’s scores was by chance.
Probability
What is probability?
the likelihood or chance that something will happen.
What is the probability equation?
P = number of actual outcomes/ number of possible oustcomes.
How do we decide if an observed difference is large enough to allow us to reject the null hypothesis?
If p = < .05
If the difference occurred by chance less than 1 in ___ e.g. 0.05 or 5%
20
What does the 5% chance correspond to?
The 5% of scores outwith the 95% - 2.5% on each side of the normal distribution
If P = > 0.05 you ____ the NH.
accept
If p= >0.05 is it significant?
no
If p = <0.05 you _____ the NH
reject
and accept the AH!
If p = <0.05 is it significant?
YES
If p is low then the null must….
GO
non-significant results don’t mean _____ results.
insignificant
What is an Effect Size?
Effect Size tells you about the magnitude of an effect rather than simply whether it is significant.
Effect sizes don’t depend on _______
sample size
Effect sizes are ____ across studies.
Standardised
effect sizes means you can _____ across studies.
compare
Null hypothesis testing is based on _____ not certainties.
probabilities
What is a Type I error?
Rejecting the NH when you shouldn’t have.
Aka. Saying something is significant when it’s not.
For controlling a type I error what is the alpha usually set at?
The alpha is set at .05.
What does setting the alpha at .05% do?
this means there is a 5% chance of making a type 1 error.
how can you decrease your chances of making a type I error even more?
Make the alpha more stringent e.g. .01
What’s the problem with making alpha more stringent?
It increases the chance of making a type II error.
What is a Type II error?
Accepting the NH when you shouldn’t have.
Aka. saying there is no effect when there one!
What is the probability of making a Type I error called?
Alpha
What is the probability of making a Type II error called?
Beta
Name the 3 factors Beta is influenced by?
- Alpha - the more stringent the alpha, the more difficult it gets to reject the NH.
- sample size - a larger sample size means greater power and reduces the likelihood of a Type II error/beta
- Effect size- a larger effect size means greater power and reduces the likelihood of a Type II error/beta
What is statistical power?
A statistical test’s ability to detect an effect if there is one.
Power is defined as 1 - _____
beta
power is measured on a _ - _ scale.
0 to 1
The closer to 1 the _____ the power.
greater
The close beta is to 1 the ____ the chance of making a type II error.
greater
Beta is measured on a __-__ scale.
0 to 1
We want beta to be very ____
small.
What is power usually set at?
0.8
What does power being set at 0.8 mean?
Your study has an 80% chance of correctly rejecting the null hypothesis. And has 20% chance of making a type II error.
What is alpha and power set at in G power?
- 05
0. 80