Ch. 11: Independent Samples t-test
How do you include effect size in your report? (ex: if ES was 1.79)
'The effect size with watching Sesame Street was very large (ES = 1.79)
How do you find the degrees of freedom (df) for the independent t-test?
(n₁ - 1) + (n₂ - 1) or N-2 (N = total sample size)
What is Cohen's guidelines for interpreting effect sizes?
0 - 0.20 = small effect size mean difference is 0 - 0.2 standard deviations 0.20 - 0.50 = medium effect size mean difference is 0.20 - 0.50 standard deviations 0.50+ = large effect size mean difference is more than 0.5 standard deviations
How do you know whether to use 1-tailed test or 2-tailed test?
1-tailed test: directional hypothesis 2-tailed test: non-directional hypothesis BUT always use 2-tailed in class
What do components of effect size formula mean?
ES: effect size x̅₁ = mean for group 1 x̅₂ = mean for group 2 s²₁ = variance of group 1 s²₂ = variance of group 2
What does statistical power influence? What doesn't it influence?
Influences likelihood of rejecting null hypothesis (i.e. statistical significance) Doesn't influence effect size
Explain independent samples t-test in terms of population and method.
Population A taught by method A µ=? ==> Sample A Population B taught by method B µ=? ==> Sample B
What is statistical power?
Power of test probability that a statistical test will correctly reject a false null hypothesis.
During exam what should you include (generically) in report?
Results for both hypothesis test and effect size
What is the amount of overlap between the 2 distributions (effect sizes)?
Small effect sizes: groups overlap a lot Big effect sizes: groups overlap very little
When reporting results in a journal article, what components are necessary to include?
Test type: t-test for independent samples Results revealed ... significant difference in ... , t(18) = 4.00, p < 0.05. Include each group with mean and standard deviation.
Where do degrees of freedom come from?
The number of scores in a sample that are free to vary.
When do you use an independent samples t-test?
Use when comparing two groups of scores from two entirely separate groups of people
How do you find the obtained value?
Using the T formula (really long)
What is an independent samples t-test also called?
between-subjects t-test
What are effect sizes?
measure of how different two groups are from on another independent of sample size related directly to magnitude (size) of difference measure of mean difference in terms of standard deviation
How do you report what you found in your independent t-test? Whether or not to reject null hypothesis in APA format.
t(18) = 4.00, p < 0.05 t: kind of test being used 18: df 4.00: obtained value p < 0.05: whether or not null was rejected
Where does the t-test use data from?
two separate groups and compares the difference between them
Know formula for calculating T.
x̅₁ = mean of group 1 x̅₂ = mean of group 2 n₁ = sample size of group 1 n₂ = sample size of group 2 s²₁ = variance of group 1 s²₂ = variance of group 2
