**Time:**Mar 3rd - 5th, 2019.- You will be given
**50**minutes to complete the quiz.

- You will be given
**Location:**Computer-Based Testing Facility (CBTF)**Quiz Scheduler:**CBTF Scheduler.

**Conflicts:** There will be *no* conflict quiz as students are able to choose the time and date of their quiz.

**All quizzes are cumulative. Previous material can reappear on a later quiz.**

5.3: Loop-the-Loops

- Iteration:
- Why are computers good at iteration and humans aren’t?
- What kind of iteration structures exist in
*R*? - How does vectorization compare to an iteration structure?

`while`

loops:- What happens in a
`while`

loop if the condition is always evaluating to`TRUE`

? - How are
`while`

loops useful to estimate numerically a solution? - What issues arise with
`numeric`

numbers?

- What happens in a
`for`

loops:- How do
`for`

loops use sequences to process data? - What are some drawbacks to supplying indices in a
`for`

loop? - Why can you write a
`for`

loop as a`while`

loop but not vice versa? - How does a
`for`

loop select the data for a given iteration?

- How do
`repeat`

loops:- Why is a
`repeat`

loop often referred to as a “do while” or running the computation at least once? - Why does the
`repeat`

loop require`break`

to exit it? - Why can’t you always write a
`while`

loop as a`repeat`

loop? - When should
`repeat`

be used over`while`

?

- Why is a

6.1: Randomness

- Random Variables
- What is a sample space? What kinds of sample spaces are available?
- What kinds of randomness exist? Which is the most common type of randomness used? Why?
- How does a
*seed*influence a PRNG? - What role does
*modulus*play inside a PRNG?

- Sampling
- How does sampling
**with**replacement differ from sampling**without**replacement? - What are some scenarios where these sampling techniques would be appropriate?
- What happens if the sample size is small compared to when it is large with respect to the stated frequencies?

- How does sampling
- Probability Distributions
- How is a probability distribution related to sampling?
- What are the different kinds of distribution functions available in
*R*? - What is the relationship between quantile and probability functions?
- Why does the
`d*()`

set of functions make little sense in a continuous sample space?

- Caches
- Why is a cache useful for simulations?
- What are some potential drawbacks to using a cache?
- How can we ensure our caches are created correctly?

6.2: Linear Regression

- Matrices
- Why are matrices useful?
- What happens during matrix construction when
`byrow = TRUE`

and`byrow = FALSE`

? - How are variables or observations added to a matrix?
- Why is matrix multiplication defined differently when compared to regular multiplication?

- Linear Relationships
- Why is a linear relationship important?
- What the different viewpoints related to modeling data?

- SLR
- Why do we perform regressions?
- How do we estimate the \(\mathbf{\beta}\) parameters?
- What are the scalar-form and matrix-form equations for linear regression?
- How does an optimizer work with an objective function?

- MLR
- How are Simple Linear Regression (SLR) and Multiple Linear Regression (MLR) related? Where do they differ?
- What are the analytical solution to Simple & Multiple Linear Regression?

- Factors
- What are factors? How do factors differ from a character vector?
- Why do we prefer storing data as a factor instead of a character vector when applicable?
- Where in statistical modeling do we find factors being useful?
- How can we coerce factors to an atomic vector?
- Why is
`stringsAsFactors = TRUE`

by default? - What is the meaning of an
*ordered*factor?

7.1: Group Work

- How many stages are there in the group project?
- Is it possible to be fired from a group for bad performance?
- What are the
*different*stages of how groups are formed? - What are some tools to help faciliate collabortion?

7.2: Bootstrapping

- Theory Overview:
- When should bootstrapping be used?
- What are the different types of bootstrapping?
- How does the underlying assumptions change between types?
- Why does bootstrap rely upon resampling?
- How can we obtain a quantile from the data?
- Where is
`1-alpha/2`

used compared to`1-alpha`

? Why is one quantity divided by 2 while the other is not?

- Implementation:
- Why does bootstrap rely upon a quantile CI?
- What is the strategy for implementing each type of bootstrap?
- Why is it better to allocate a vector of
`NA`

values instead of directly specifying a specific data type to hold the bootstrapped statistics?

7.3: Hypothesis Testing

- Testing Frameworks:
- Why do we use hypothesis testing?
- What is a sampling distribution and how are they used in hypothesis testing?
- What are the four stages of a hypothesis test?

- Ides of Testing:
- How do research questions inform the
**null hypothesis**and**alternative hypothesis**? - What are the kinds of assumptions are required to use a hypothesis test?
- Does having more numbers after the decimal place in a test statistic guarantee more accurate results?

- How do research questions inform the
- Unpaired (Two-sample) t-Test:
- Is a numeric number more “trustworthy” if there lots of digits after a decimal point?
- How is a test statistic used when computing a
*p*-value? - Do we have to compute a
*p*-value to make a decision? - What are some misconceptions about
*p*-values?

- One proportion z-test:
- What happens if a probability value needs to be assessed?
- How does the underlying assumptions between probability and samples differ?

Students will have access to:

*R**RStudio*- Any
*R*package covered up to this point. - The following
*RStudio*cheatsheets: - Casio FX-300MS Scientific Calculator
- Notes on how to access
*RStudio*. - Video showing how to access
*RStudio*.

- Preferably, a
**rested mind**and non-broken hands that can type.

- All answers must be reasonably simplified.
- Decimals answers must contain
*two*significant digits. - Grading will be done as follows:
- A correct answer will receive all points.
- An incorrect answer will receive proportionally appropriate partial credit.

If you have a technical issue while answering questions or need assistance with opening or starting the quiz, please alert the proctor.

Do *not* leave the CBTF without filing an issue with the **proctor** if something goes wrong.

Have a testing accommodation? Please see how the CBTF handles Letters of Accommodation.

The short version: **Please bring a copy of the Letter of Accommodation to the CBTF Proctors prior to the test taking place.**

In short, don’t cheat. Keep your eyes on your own quiz. Do *not* discuss the quiz with your friends after you have taken it. Any violation will be punished as harshly as possible.

The best way to *study* for a STAT 385 quiz is by *writing* and *reading* code. Try to take an idea in STAT 385 and apply it to your *own* work.

With this being said, there are *three* other resources that may assist your studies:

- Topic Outline (Above)
- Lecture Code
- Homework

Again, the best way to study is to *do* programming in some fashion. Whether that be *writing* code or *explaining* how code works to someone else.

Consider using resources such as:

- RStudio Cloud Primers for interactive practice.
- Exercise problems listed in a given section of the readings.

Do *not* spend time memorizing lecture slides. You will *not* see any verbatim questions.

Do *not* try pulling an all-nighter. You can *schedule* your quiz anytime between a time window. To program efficiently, you need sleep despite the quote:

“Programmers are an organism that turns caffeine into code.”

There are generally four types of problems:

- True / False
- Multiple Selection (e.g. select ALL correct answers from a list)
- Fill in the blank
- Writing Code

Only one question with 15012391 subquestions. In all seriousness, do not fixate on a number. There will be a reasonable amount of questions for the time period.

Depending on your background, the quiz may take:

- Prior
*R*in-depth experience: 25 minutes - Some
*R*experience: 35 minutes - No
*R*experience: 50 minutes

Avoid fixating on *time*. Life will come and go more quickly than you realize. Focus more on the content.

As all problems are automatically graded, we should be able to post the quiz results after the examination window closes.

No.

No. Curving is only done sparingly at the *end* of the semester. Individual assignments are not modified.