Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
27 Cards in this Set
- Front
- Back
Robust means that:
|
When the confidence level or P-value does not change very much even when the conditions are not met.
|
|
To be Robust, what are the conditions?
|
No outlier, no extreme skewness.
|
|
a= ?
|
a= symbol for level of significance
|
|
Purpose of confidence interval is to estimate the:
a. Population mean b. Sample mean |
a. Population mean
|
|
True mean is also known as:
|
Population mean
|
|
If asked what is the appropriate formula for a confidence interval for mu what do we look for?
|
1. If the population size is < 30 we use t* , if >30 we use z*
2. When population st dev is not given - use t* (First determine: Population size? Second: is population stdev given?) |
|
You can only use t* distribution when the distribution is:
|
normal
|
|
When n=less < 30 make sure ther are no:
|
outliers or skewedness
|
|
If asked "What would be checked in order to validly use given formula when n=15" ....what do we look for?
|
Make sure there is no outliers or strong skewedness in a plot of the data. (when n=<30)
|
|
If the sample size is the same and a= same and o= same for both groups then the margin of error are:
|
the same
|
|
When n increases, the margin of error:
|
decreases
|
|
When CONFIDENCE LEVEL increases, the margin of error:
|
Also increases
|
|
Can you recommend using a one-sample t confidence interval estimate for mu if there is an outlier in the data?
|
No, t distribution is not appropriate in this case since there is an outlier.
|
|
Rule for using confidence interval of the difference between two means:
(Reject, Fail to Reject?) |
Reject Ho if the confidence interval does NOT contain zero.
Fail to reject Ho if the confidence interval does contain zero. |
|
Reject Ho if the confidence interval:
|
does NOT contain zero
|
|
Fail to reject Ho if the confidence interval:
|
does contain zero.
|
|
If asked to conclude that there is a significant difference aka reject/fail to reject Ho, you must look for what words in the problem?
|
1. For the difference
2. Look at the mean intervals (does it contain a zero?) 3. significant difference |
|
If asked if we can conclude the mean variable differs from a given data variable what do we look for?
Sample Question: You want to compare the daily items sold for two game consoles: Playstation3(PS3) and NintendoWII(WII). Over the next 80 days, 40 days are randomly assigned to PS3 and 40 days to WII. At the end, you compute a 95% confidence interval for the difference in mean daily items sold for the two game consoles to be (-20, -10). On the basis of this confidence interval, can you conclude that there is a significant difference between the mean daily items sold for the two game consoles at α=0.05? (i.e., can you reject Ho?) |
Look between the confidence interval (2.84,3.06) and determine if the given data variable (3.1) would show up between the two confidence interval data points
|
|
Margin of error questions use the n=z*o/m^2 formula...when you finish you always round:
|
up
|
|
Which of the following questions does a test of significance answer?
|
Is the observed effect too large to be due to chance alone?
|
|
Which one of the following statements best describes the logic of tests of significance?
|
An outcome that would rarely happen if Ho were true is good evidence that Ho is not true.
|
|
Type I error and Type II errors are:
|
Type I - when a true null hypothesis is rejected
Type II - when a false null hypothesis is not rejected (fail to reject Ho when Ho is false) |
|
Increasing the sample size increases the _____ of the test.
|
power
|
|
The P-value for a significance test is defined as
|
the probability of obtaining a test statistic that has a value at least as extreme as that actually observed, assuming the null hypothesis is true.
|
|
Small P-value means that there is stronger evidence AGAINST the
|
Null Hypothesis
|
|
Large P-value SUPPORTS the
|
Null Hypothesis
|
|
What does the normal curve represent if a one-sided test on Mu with o known, and the p-value represented as the area in the tail of the normal curve.
|
All possible values of and how often they occur if Ho were true.
|