Midterm 2 Outline

Possible questions

Still under construction

Plots

Create a box and whisker plot.

Interpret a normal quantile-quantile plot.

Maximum likelihood estimator derivation.

Poisson

Exponential

Bernoulli trial

Geometric

Maximum likelihood estimate of a parameter from data

example: estimate lambda from data an exponential distribution based on sample data

note that this will be a number, not the derivation.

Confidence intervals

We have the list

Tests of parameters

We have the list

assumptions, hypotheses, test statistic, how is it distributed, etc.

Test from a multinomial distribution

There are a few variants, e.g. test of independence in a joint table

assumptions, hypotheses, etc.

There will be some study questions, from this set:

SQs

Study Questions, Class 9

1. What is a random sample?

2. What is a statistic?

3. What are two reasons why we might want to summarize a sample with a statistic?

4. A sampling distribution is the probability distribution for a _____________?

5. What do we look for in a normal quantile-quantile plot that indicates the sample comes from a normal distribution?

---------------------------------------------------------------------------

Study Questions, Class 10

3. Compare the shape of the T distribution to that of the Normal distribution:

---------------------------------------------------------------------------

Study Questions, Class 11

1. What is it that we find estimates for, using a random sample?

2. What is the difference between an estimator and an estimate?

3. What special symbol do we add to our Greek letter names to indicate estimators and estimates?

6. Define unbiased.

7. Define efficient.

8. Define consistent.

---------------------------------------------------------------------------

Study Questions, Class 12

1. How is the model for paired observations different from all the other sampling models we have been looking at?

---------------------------------------------------------------------------

Study Questions, Lecture 13

1. A statistical hypothesis is a statement about what?

3. We often identify a null hypothesis and an alternative hypothesis. What does a null hypothesis have to be that an alternative hypothesis doesn't have to be?

5. What is the relationship between alpha and the significance level?

7. What is the relationship between power and the sensitivity of a test?

8. Which hypothesis do we look to calculate beta and power? What form does that hypothesis have to have in order to do the calculations?

9. How can we increase the power of a test, even if we can't calculate it?

---------------------------------------------------------------------------

Study Questions, Lecture 14

5. What should be true about the parent distribution for a paired t-test, in order for it to have more power than a two-sample test on the same numbers?

---------------------------------------------------------------------------

Study Questions, Lecture 15

nada

---------------------------------------------------------------------------

Class Home