Study Guides
(238,266)

United States
(119,714)

New York University
(633)

Math
(12)

MATH-UA 235
(1)

James P.Stanley
(1)

Final

# AP Statistics- Final Review.pdf

Unlock Document

New York University

Math

MATH-UA 235

James P.Stanley

Fall

Description

AP Statistics: Final
Test 7
I. Forms of Studies/ Sampling
A. Simple Random Sample (SRS): each subject is equally likely to be selected; each sample
of size is equally likely to be selected
B. Stratiﬁed Random Sample: divide the population into strata (plural); from each stratum
(singular), select an SRS
C. Proportionate Allocation: the sample size is partitioned among the strata proportional to
the size of the strata
D. Stratiﬁed Random Sampling: population is divided into homogeneous (same) groups. An
SRS is them selected from each group. (Note: Groups are heterogeneous with respect to
each other.)
E. Systematic Random Sampling: a random sample is sampled, then every kth subject after
him/her is sampled
F. Cluster Sampling: separate the population into heterogeneous groups. With respect to
each other, the groups are homogeneous
G. Convenience Sampling: Conducted in 1 location to obtain the responses of a convenient
group. Problem: Sample does not always represent the population
H. Multistage Sampling: combines different types of sampling
II. Bias
A. Selection Bias: the sample is not representative of the population
B. Voluntary Response Bias: people volunteer to respond causing the sample to be non-
random
C. Measurement Bias: the instruments used for recording data are not accurate
D. Response Bias: anything in a survey that inﬂuences responses
E. Nonresponse Bias: people refuse to respond or can’t be reached
III. Studies
A. Observational Study: we observe people who are selecting their own habits and we
collect data on the variable(s) of interest
B. Experimental Study: treatments are imposed on subjects. The goal is to attribute cause/
effect
C. Retrospective Study: an observational study in which subjects are selected and then their
previous conditions or behaviors are determined
D. Prospective Study: an observational study in which subjects are followed to observe
future outcomes
IV. Experimental Design
A. Experimental Units: that which we are applying a treatment to (when experimental units
are people, we call them subjects)
B. Explanatory Variable (Factor): whatever we expect to affect outcomes in the study (there
can be more than 1 in a study)
C. Response Variable: the variable whose values are compared after treatment
D. Levels of the Factor: values that factor(s) can take on
E. Treatments: if 1 factor, treatment is same as factor; if more than 1 factor, all combinations
of levels of the factors
F. Diagram of Experimental Studies (in order) SUBJECT --> (RANDOM ALLOCATION) --> GROUPS --> TREATMENT --> MEASURE
▯ ▯ ▯ ▯ ▯ ▯ ▯ ▯ ▯ ▯ ▯ RESULTS
G. Principles of Experimental Design
1. Randomization
a) Goal is to equalize the groups with respect to variables that may impact the
response
2. Control
a) Control sources of variation
3. Blocking
a) When groups of subjects are similar, gather them into blocks (reduces variability in
results)
4. Replication
a) Repeat the experiment using different subjects OR use multiple subjects in each
treatment group
H. Matched Pairs Design: participants in different conditions are matched to certain
characteristics
V. Mean and Standard Deviation
A. Use x for sample mean which is a part of a population
B. Use µ for population mean which is the mean of all sample means
x
C. Use S xfor the sample standard deviation
D. Use σ xfor the population standard deviation
VI. Sampling Distribution of Sample Means
A. We select an SRS of size n and record mean ( x ) for the variable of interest
B. Continue selecting SRS’s of size n and recording x . There are N C n possible SRS’s ( N
< population size)
C. The distribution of all means is the sampling distribution of sample means
VII. Properties of the Sampling Distribution of Sample Means
A. The mean of the sampling distribution of sample means is equal to the population mean
B. The standard deviation of the sampling distribution of sample means is found by
σ = σ x where σ is the population standard deviation and n is the sample size
x n x
C. In most cases, the sampling distribution of sample means is roughly Normal.
VIII. Sample Size and Binomial Model
A. When a parent population is Normal, the sampling distribution of sample means is Normal
B. If the sample size ≥ 30, the sampling distribution of sampling means is Normal
C. The binomial model is approximately Normal when np ≥ 10 and nq ≥ 10
Test 8
IX.Conﬁdence
A. How to Find the Conﬁdence Interval
* *
1. The conﬁdence interval is found by p±margin of error or p±(z ×se(p))ˆ where z is
the z score for the conﬁdence level and se(p)is the standard error
2. The z score is found by marking the percentage of the conﬁdence level as the center
of a Normal distribution with half of the conﬁdence level to the left of the center and the
other half to the right of the center; then, ﬁnd the area left or right of the marked off
point; then ﬁnd the z score of that and use the absolute value of it pq
3. The standard error is , found by se(p) =
n
4. The conﬁdence interval is reported in parenthesis, where the lower value is followed
by a comma and the higher value
B. Margin of Error
1. The margin of error decreases when the sample size increases
2. If we want to keep our conﬁdence level but reduce the size of our conﬁdence interval,
we need to make the margin of error smaller
3. Smaller samples will produce larger margin of errors
C. Standard error is the same as standard deviation but it takes into account of the sample
size.
D. Standard error decreases as sample size increases
E. Reasonable Variability: the amount of variability we expect to see between the different
samples of the same size
X. One- Proportion Z-Test
A. Assumptions and Conditions
1. Was the selection done randomly?
2. Is each selection independent of each other?
3. np ≥10;nq ≥10
o o
4. Is the population more than 10 times the sample?
B. Alpha Value
1. Lower alpha value --> rejection region is lower
2. Higher alpha value --> rejection region is higher
3. If p-value < or = to alpha value, reject the null hypothesis
4. If p-value > alpha value, fail to reject the null hypothesis
C. How to do a 1- Proportion Z-Test
1. Set up null and alternate hypotheses
2. Verify conditions and assumptions
3. Find p
4. Calculate zˆp
5. Find the p-value and interpret it
D. A p-value is statistically signiﬁcant if the null hypothesis is rejected
E. In a two- tailed hypothesis test, multiple the p-value by 2 because both extremes are
being accounted for
Test 9
I. Conﬁdence Intervals
A. Conﬁdence Level = 100(1-α)%
B. If 0 is not captured by a 100(1-α)% conﬁdence level (p is no0 in the range of the
interval), reject 0
C. If p0is captured by a 100(1-α)% conﬁdence level (p is in0the range of the interval), fail to
reject H0
II. Errors and Power
A. Type I Error: made when H was0true but was rejected
B. Type II Error: made when H was0false but was failed to be rejected
C. Power: made when H is f0lse and was rejected
D. The correct decision occurs when H was 0rue and failed to be rejected E. The probability of a Type I error is α
F. The probability of a Type II error is β
G. α and β do not add up to 1!
H. The power is 1- β
I. Effect Size: the distance between P an0 P (null and true)
J. All other things being held equal, as effect size increases, the probability of a type II error
decreases, the probability of a type I error increases and the power increases
K. If we want to decrease the probability of a type I error, we can decrease α
L. If we want to decrease the probability of a type II error, we can increase effect size or
sample size
III. 2-Proportion
A. Assumptions and Conditions for Inference about a Proportion
1. Both data are from an SRS from the population of interest
2. Both populations are at least 10 times as large as the samples
3. n is so large that n1p1, n1(1− p1), n2p2, n2(1− p 2 are 5 or more
IV. Hypothesis Testing for Means
A. Assumptions and Conditions
1. Randomization Condition OR The sample is representative of the population
2. Independent Condition
3. 10% Condition
4. Normal or nearly Normal Condition
B. Because this is concerned with means and not proportions, the mean for testing is µ x,
σ x
the mean of the sample is x and the standard deviation of the sample is

More
Less
Related notes for MATH-UA 235