Experimental Design
Sampling
Regression
Statistical Tests
Hypothesis Testing
100
The simplest true experimental design
What is Random Sample Experiment
100
This is why we sample (at least one of the reasons)
What is cost, time, feasibility, representativeness?
100
This is ALWAYS the first step in modeling the relationship in bivariate data (after it is collected).
What is "eyeball the data?"
100
This test is appropriate for nominal data with only two values, and when we know the long-haul probability of these values appearing.
What is the Binomial Test?
100
This is the logical converse of the hypothesis we really WANT to test
What is the null hypothesis?
200
This is the unbiased method of dividing a sample into Experimental and Control conditions
What is Random Assignment
200
We use these to estimate population parameters
What are sample statistics?
200
This is an appropriate thing to do when you find a distinct pattern in your residual analysis
What is transform the data?
200
This test addresses the null hypothesis that the proportion(s) of data in a contingency table are all equal.

What is the X2 Test of Homogeneity?

200
This is claiming that a null hypothesis is false, when in fact it is true
What is a Type I error?
300
This is the chief advantage of a Random Sample Design over an Ex Post Facto Design
What is, Random Sample Design makes the Experimental And Control Groups functionally equivalent, so that the only plausible difference between them is just random chance?
300
A distribution of sample statistics all drawn from the same population, with the same n
What is a Sampling Distribution?
300
This tests to see if the slope of your regression model is significantly different from zero

What is t-test of the slope or t-test of B1?

300
This is the ONLY conceptual difference between the one-sample z-test and the one-sample t-test
What is, for the z-test, you know the population sigma, while for the t-test you do not know the population sigma and have to use the sample standard deviation instead.
300
Power is inversely proportional to this error
What is Type II error?
400
This is one of the advantages of a Pre-Post design
What is Cost, Simplicity, Ability to attribute difference to treatment?
400
The first tells us a single sample is better with larger n, the second tells us that as n grows, the width of the sampling distribution reduces.
What are, The Law of Large Numbers, and Central Limit Theorem, respectively.
400

This is the relationship between SSregression, SStotal, and SSresidual

What is SSregression +  SSresidual = SStotal

400
The assumptions for this test are:

1. Random sampling from a CRV distributed as Normal;

2. Homogeneity of variance;

3. Independence of Observations

What is the two-sample t-test?
400
We do this when the magnitude of a test statistic is greater than the critical value of that statistic
What is reject the null hypothesis?
500
This design reduces nuisance variation in the data by first dividing the population into pre-existing groups, then randomly sampling from those groups into experimental conditions
What is Randomized Block Design?
500
This makes any bias in the representativeness of a sample solely due to chance
What is Random Sampling?
500

This is the definition of R2

What is, The proportion of variability accounted for in the regression model?


also acceptable, "SSregression/SStotal

500
This test uses a pre-post experimental design, with a CRV as the dependent measure, testing the hypothesis that there is no mean difference in performance from pre- to post-
What is the Matched Pairs t-test?
500

This is the region that 1 - alpha% of the samples of size n, all taken from the same population, will likely fall in the sampling distribution if the experiment were repeated infinitely many times.

What is 1-alpha confidence interval
M
e
n
u