Comparing Means Flashcards
What is the benefit of comparing means?
Allows us to see if a change in an IV leads to a change in a DV
What is a null hypothesis?
Expect no difference between means
What is an alternative hypothesis?
Mean of one group different from mean of other
What are descriptive statistics?
Numbers describing the sample
What are inferential statistics?
Using descriptive statistics to find out about the population
What is the problem with using samples?
Data from two samples will give different means, even if they come from same population
What is a p-value?
How well the data from the sample fits the null hypothesis
Probability of seeing difference in the means in the sample given there was no difference in the means in the population
What is statistical significance?
If this p-value low, take as a “sign” that there is a real difference in the population
Can be confident that the means are different in the population if p<.05 (less than 5%)
If p>.05 then non-significant
What is a t-test?
In population, different samples give different means
If we take means from many samples, can plot means as a sampling distribution
More standard errors the means are away from each other, the more confidence that the means come from different distributions
What does the standard deviation of a sampling distribution depend ono?
Sample size (per sample)
Standard deviation (of the variable)
What does a t-test tell us?
Student’s t-statistic
P-value
What is a student’s t-statistic?
How many standard errors the means are away from each other
What are degrees of freedom?
Before calculating p-value, must calculate degrees of freedom
Number of parameters (numbers) which are free to vary
Amount of information (numbers) we know - the more information the confident we can be in our results
What is the equation for degrees of freedom?
df = n - 1
What is an independent samples t-test?
Used for between-subjects design
Have 2 groups and want to see if they have significantly different means?
What are other names for an independent samples t-test?
Unpaired means t-test
Unrelated samples t-test
Between groups t-test
What are the assumptions of an independent samples t-test?
Interval/ratio data
Data is independent
Variables are normally distributed
Both groups have similar variances (therefore similar standard deviation)
How do you test if groups have similar variances?
Levene’s test - tests for equality of variance
What does it mean in the Levene’s test if p >= .05>
Assumption of equal variance met
What does it mean in the Levene’s test if p < .05?
Assumption of equal variance not met
What is a paired-samples t-test?
Used for within subjects design
Measures people at 2 different times
What are the other names for a paired-samples t-test?
Related-samples t-test
Paired means t-test
Within subjects t-test
Repeated-measures t-test
How powerful is a paired-samples t-test?
More powerful than independent sample t-test
More likely to find true effect
Less error (as measuring same people)
How is the t-statistic calculated in a paired-samples t-test?
How many standard errors is the mean different from zero
Numerator = what is the mean difference
Denominator = standard error
How is statistical significant calculated in a paired-samples t-test?
More standard errors the mean difference is from zero, the more confident we can be that there is a real difference
p < .05 if t-statistic is in critical region
What are the assumptions of a paired-samples t-test?
Interval/ratio data
Random sample from population
Difference between scores is normally distributed
What is a one-sample t-test?
Compare mean to test value (X)
