Upgrade to remove ads
Only $35.99/year

Analytics Exam 2

Terms in this set (13)

Confidence Interval Formula
Upper Bound: sample mean + confidence interval decimal
Lower Bound: sample mean - confidence interval decimal
Effect Size Formula
(sample mean - population mean) / standard deviation

Z-Test: scenario STDEV
T-Test: data STDEV
Standard Error Formula
(STDEV) / True Sample Size ^ 0.5
the 4 assumptions for a T-Test
1. dependent variable is scale - ratio or interval (absolute zeros)
2. observations are randomly selected - found in scenario
3. population is normally distributed - 30 or more required for normal distribution
4. we should have a population mean - found in scenario
the 4 assumptions for a Z-Test
1. dependent variable is scale - ratio or interval (absolute zeros)
2. observations are randomly selected - found in scenario
3. population is normally distributed - 30 or more required for normal distribution
4. parameters for 1 population (is there a mean AND a standard deviation provided)
How to determine a 2 tail test from a 1 tail test
2 tail is NOT based on previous research so we do not know what to expect - it could go either way.
1 tail is based on previous research so we do know what to expect - we already know which way it goes.
Hypothesis for a 2 tail test
Null: there is no difference
Research: there is a difference
hypothesis for a 1 tail test (Scenario- "has been shown to increase..."
Null: lower or equal to
Research: higher
hypothesis for a 1 tail test (Scenario- "has been shown to decrease..."
Null: higher or equal to
Research: lower
Do we get the descriptive statistics before or after we double the date
Always BEFORE
The effect size explanation
we would expect that any individual ... will have a .... by .... standard deviation than someone .... This is a .... effect size.
The confidence interval explanation
There is a ....% chance that the true population mean for .... any individual ..... between ... and ....
If you wanted to increase the statistical power for your research study, what would be the best way to increase it?
Increasing the sample size. This is the best choice because we increase power without effecting the alpha and the tails.
Other sets by this creator