PSYC-2019 Quiz 3
What type of t-test compares a sample mean and a population mean when population SD is not known?
Single Sample
What t-test compares two samples and every participant is in both groups?
Paired-samples
What t-test compares two samples and every participant is in only one sample?
Independant samples
If a study mentions a 'before and after' what t-test should be used?
Paired samples
When is a t distribution used? (2 points)
When no population SD is provided
There is more than one sample
what happens to a t distribution when there are less subjects?
It is harder to find signifigance
the z distribution becomes wider
What formula is this for: N-1?
Degrees of Freedom
What must we be careful NOT to do when a sample is not 100% random?
Over generalize
What is a standard minimum sample size that we need to generalize to a population?
30
What are you finding when asked to determine the characteristics of the comparison distribution?
Standard Error
When should you use the standard error of a population?
When you don't know the SD
How are critical values/cut offs determined?
Using a provided tabled with the degrees of freedom and alpha
For a two tailed test, is the critical value positive or negative?
It can be both
As your critical values get further from the mean what happens to the t and p values?
t value gets larger
p value gets smaller
What does it mean to 'make a decison'?
Determine what is bigger, the t-obtained or the t-critical
If the t-obtained is larger than the critical value, what can we say about the null?
We can reject the null
If the t-obtained is larger than the t-critical what does this say about signifigance?
It is signifigant
If the t-critical is larger than the t-obtained, what does that say about the null?
We fail to reject the null
What is t-obtained?
The test statistic
What number is a small effect size?
0.20
What number is a medium effect size?
.50
What is a large effect size?
0.80
Which t-test involves each participant being assigned to only of the sample groups?
Independant-samples
What does a mean difference of zero indicate?
No difference between the samples
Which t-test uses between-subject designs?
Independant-samples
What distribution is used for an independant-samples t test?
Distribution between means
What distribution is used for a paired samples t-test?
Distribution of mean difference scores
What distribution is used for a single sample t-test?
Distribution of means
What are the 4 assumptions of an independant-samples t-test?
1) DV is a scale variable
2) Random samples are used
3) Population is normal
4) There is equal variance
If there is not quite equal variance in an indepedent-samples t-test, what is the general rule for using it?
It can be used as long as one variance is not 4 times larger than the other variance
For an independent-samples t test do you need an equal number of subjects in each group?
No
What does it mean if your p-value is bigger?
it is bad because the probability that the null is correct is higher