Stats exam 3 study guide-
5 questions on estimation :
Don’t need to calculate confidence intervals
How to tell if there is a sig difference based on confidence intervals:
Comparing 2 groups; 95% interval [10,15] is there a difference? Yes, one group is bigger by 10-15 points
Comparing 2 groups; 95% interval [-2,20] no, because 0 is included
[-2,-15], yes one group is always less than the other groups
Tested a sample of teenagers and 95 percent sure the sample of n=100 is between [6,8]
Formula for confidence intervals applies to all everyone not just your sample n
Mean score of population
Point estimate
5 people class A mean=50
5 people class B mean= 60
Point estimate=10 points
10 students give exam mean=70
10 students give exam after coffee mean=60
Point estimate=-10
Don’t know confidence but it is the guess
For all statistics:
Effect size: cannot influence effect size, can influence ability to detect it—the larger the sample size the
more ability to detect it (power)
Variability- makes less likely to find significance and decreases power
Given example of difference groups and their means—as soon as its more than 2 groups its an anova, 2
groups compared 2 means it is a T-test—people were tested and then tested again (pre, post) it is
5 questions on estimation :
Don’t need to calculate confidence intervals
How to tell if there is a sig difference based on confidence intervals:
Comparing 2 groups; 95% interval [10,15] is there a difference? Yes, one group is bigger by 10-15 points
Comparing 2 groups; 95% interval [-2,20] no, because 0 is included
[-2,-15], yes one group is always less than the other groups
Tested a sample of teenagers and 95 percent sure the sample of n=100 is between [6,8]
Formula for confidence intervals applies to all everyone not just your sample n
Mean score of population
Point estimate
5 people class A mean=50
5 people class B mean= 60
Point estimate=10 points
10 students give exam mean=70
10 students give exam after coffee mean=60
Point estimate=-10
Don’t know confidence but it is the guess
For all statistics:
Effect size: cannot influence effect size, can influence ability to detect it—the larger the sample size the
more ability to detect it (power)
Variability- makes less likely to find significance and decreases power
Given example of difference groups and their means—as soon as its more than 2 groups its an anova, 2
groups compared 2 means it is a T-test—people were tested and then tested again (pre, post) it is
