But I Regress...
The Column Without A Cool Name
I've got 99 Probabilities and Significant Ain't One
What's With All the Hyp(e)
Room of Requirement(s)
100
All linear correlation coefficients are between these two values.
Between -1 and 1.
100
All confidence intervals for either estimates of means or proportions are essentially just
_____ ±______
point estimate ± margin of error.
{Or point estimate ± t/zc*standard error}
100
What is the probability that a z score will be greater than -1.31?
.9049
100
In a test with
alpha=0.05
and p-value 0.048, you would conclude this.
Reject H0
100
A friend did a one sample z-test on the following data. Why is she wrong?
A random sample of 22 fifth grade pupils have a grade point average of 3.5 in math with a standard deviation of 0.452. The grade point average (GPA) of all fifth grade pupils of the last five years is 3.2. Is the GPA of the 22 pupils different from the populations’ GPA?
A z-test requires that we know the population standard deviation.
200
When you test for
rho > 0
you are also testing for this.
Test for a positive slope
beta >0
200
What happens to a confidence interval as n gets larger
The confidence interval gets smaller.
because the margin of error is smaller
the confidence interval gets more precise.
200
The z-score that corresponds to the top 10% of all scores.
z=1.28
200
The formula for calculating a t-value
t= (bar x- mu)/(s/ sqrt(n))
200
The requirement to use a t-test if the data might have a slight skew.
n should be greater or equal to 30
300
The interpretation of the slope of the equation
hat y = 3.08-6.35x
For every one unit increase of x, y decreases 6.35 units.
300
The interpretation for the 90% confidence interval for statistics class's test scores that had a mean of 81, and a margin of error for the mean of +/- 4.
We are 90% confident that the interval from 77 points to 85 points is one that contains the population mean score of the test.
300
What is the full calculator command (with parameters) for finding the probability of more than 6 successes in a binomial experiment with 10 trails when p=0.2
1-binomcdf(10, 0.2, 6)
300
The Null and Alternative Hypotheses for the experiment:
Since 2006 at the Tennis US Open, men have challenged 1412 referee calls and had 421 overturned, and women have challenged 759 and had 220 overturned. We want to test the claim that men have better success challenging calls than women.
H0 : p1 = p2
H1 : p1 > p2
300
The requirement/s for a discrete probability distribution to be considered legitimate.
The probabilities must add to 1,
and
The probabilities each must be between 0 and 1.
400
The t-value and conclusion for the test for linear correlation on the data:
x: 1 | 3 | 5
y: 7 | 10 | 9
t=0.8660
No significant linear correlation.
400
A list of at least 5 greek letters used in stats class, along with what they represent.
To Sum:
Sigma
Population standard deviation:
sigma
Population regression slope:
beta
Population regression y-int (or significance level):
alpha
Pop. Mean:
mu
Pop. Linear correlation:
rho
A distribution for variance:
chi
Degrees of freedom:
nu
400
P(1st Class and Survived)
and P(Survived | 1st Class)
0.1526 and 0.6204
400
The p-value for a test statistic that has a value of 3.1 when n=2, and your alternate hypothesis looks like
H_1: sigma^2 >12
chi^2
0.05<p-val<0.1
400
When you want to compare the means of two different samples and no population standard deviations are known, these are the four different tests that could be used, with reasons why we'd used each.
Sign-test for dependent samples, no parameters assumed
t-test for Dependent samples, if normal distribution
Rank-sum is for independent samples, no parameters assumed500
An r2 value of .87 tells you this.
That only 87% of the variation in y is explained by the regression of y on x.
500
The mean of the grouped data:
12.29
(10*8+13*5+16*4)/17
500
The probability that 25 runners have a mean resting heart rate of less than 53 bpm, Assuming all runners have a mean resting heart rate of 57 beats per minute with a standard deviation of 7 bpm.
0.0021
500
If a p-value is 0.013 for a test statistic of t=2.5, that means that 0.013 is the probability that ________________ , assuming ____________.
The probability that the t will be as extreme or more extreme than 2.5, assuming H0 is true.
500
___A____ is used for testing data in a contingency/2-way table, ____B____ if a distribution of data fits the expected, and ___C___ is for when you want to test what a t-test does but have too many samples.
A: Test for Independence/Homogeneity
B: GOF-test
C: ANOVA