Why do we need to have an SRS for inference?

Allows us to generalize to the population.

Why does a larger random sample give more trustworthy results than a smaller sample?

This will reduce variation among the samples.

The Central Limit Theorem says that when n is large,

the sampling dist'n is ≈ normal

What does blocking do in an experiment?

Accounts for variation in the response due 2 block variable

systematic variation in the data

probability of getting result this extreme or more given Ho

studbud

AP Review Q3

32 termsStudBud Administration123 plays

Quick review of commonly tested items on the AP Statistics Exam. #apstatistics #apexamreview #examreview

Flashcards32 cardsFPractice quizPHost liveL

Questions 32

Answers

1
Which of the following is NOT a required condition for a 2-sample t-test?
large enough sample sizes
independent groups
independent observations
random samples
2
Confounding typically occurs
in observational studies
when blocking is not used
in linear regression t-tests
in a randomized experiment
3
Capital of Montana
Butte
Bozeman
Helena
Billings
4
Capital of Nevada
Henderson
Carson City
Reno
Las Vegas
5
Why do we need to have an SRS for inference?
To ensure independence.
Allows us to generalize to the population.
It eliminates bias.
It eliminates confounding.
6
A sample size calculation for MOE ≤ 0.03 and 95% confidence gives n ≥ 78.2 How should we round?
round to 79
round to 78
don't round at all
7
Capital of Michigan
Grosse Pointe
Detroit
Ann Arbor
Lansing
8
Capital of Wisconsin
Madison
Green Bay
Appleton
Milwaukee
9
We use a t-model when
we know s but not sigma
the sample size is small
we know sigma but not s
the sample size is large
10
Capital of Texas
Dallas
Fort Worth
Houston
Austin
11
Capital of California
San Francisco
Sacramento
Los Angeles
San Diego
12
Free response answers should always be in
complete sentences
pen
pencil
context
13
r (correlation) should be used ONLY when
when all conditions are true
in a randomized experiment
the data comes from a SRS
the data are linear
14
Capital of Florida
Tallahassee
Orlando
Tampa
Miami
15
Why does a larger random sample give more trustworthy results than a smaller sample?
It will reduce bias.
It will decrease the standard deviation.
It will reduce variation in the sample.
This will reduce variation among the samples.
16
State capital of Missouri
Springfield
Jefferson City
St Louis
Columbia
17
In the table, E(X) ≈ 7.18 (the expected value). How should we round?
don't round at all
round to 7
round to 8
18
When the p-value is larger than alpha, we
accept the null
cannot reject the null
reject the null
need a larger sample size
19
Capital of Washington
Vancouver
Seattle
Spokane
Olympia
20
In a two-sample t-test, n1=25 and n2=30. What is the most conservative degrees of freedom?
54
29
53
24
21
Capital of Iowa
Iowa City
Ames
Des Moines
States don't have capitals.
22
The Central Limit Theorem says that when n is large,
random samples are not needed
the sampling dist'n is ≈ normal
the data is normal
the standard deviation formula can be used
23
Widgets have std. dev. = 100g. The std. dev. of the sampling distribution of x-bar when n=25:
25g
20g
100g
4g
24
Capital of New York
Rochester
Albany
New York City
Buffalo
25
% of variation in the y variable that is explained by the model on the x variable
standard deviation
r (correlation)
standard error
r-squared
26
What does blocking do in an experiment?
It eliminates bias.
It helps create equal treatment groups.
It eliminates variation.
Accounts for variation in the response due 2 block variable
27
When comparing two situations (distributions, scatterplots, etc.), always include
correct spelling
correct grammar
comparative language
lots of facts
28
Bias is
any variation that affects the results
systematic variation in the data
personal beliefs mixed in with data
naturally occurring variation in data
29
A p-value is the
probability that you made a Type I error
probability of getting result this extreme or more given Ho
probability that the null hypothesis is true
probability that the alternative hypothesis is true
30
A Type II error is
not rejecting a false Ho
rejecting a true Ho
rejecting a false Ho
not rejecting a true Ho
31
How do you decrease Type 1 error?
increase n
increase alpha
decrease alph
you can't decrease it
32
Randomly calling a household for a survey, only to have them hang up on you is
response bias
undercoverage
nonresponse bias
rude bias

AP Review Q3 Flashcards | StudBud