Stats 13 Final Exam

MIDTERM 1

YAY

population

all items of interest

sample

portion of the population

parameter

numerical summary measure about the population

statistic

summary measure about the sample

observational units

individual entities on which data are recorded

quantitative

numerical

two types of quantitative variables

discrete and continuous

discrete

countable

continuous

measurable

categorical

attribute or characteristic

non-sampling errors

any errors that result from the survey process

sampling error

the error that results from using sampling to estimate information regarding a population

examples of non-sampling errors

incomplete population, poor interviewer, poorly worded question, incorrect data entry

pi

true probability

p-value

the probability of obtaining a value at least as extreme as the observed statistic when the null hypothesis is true

type I error

occurs when you reject the null

type II error

occurs when you fail to reject the null

p hat

observed proportion

n

sample size

what best describes a sampling distribution model of a statistic?

a. it is the probability that the sampling statistic equals the parameter of interest

b. it is the probability distribution of all values that are contained in all possible sample of the same size

c. it is a distribution of all the statistics calculated from all possible samples of the same size

d. it is a histogram of sample statistics from all possible samples of the same size

c. it is a distribution of all the statistics calculated from all possible samples of the same size

what number does the z score have to be above in order for the results to be extreme

2

the graph of a null distribution will be centered approximately on the

a. observed proportion

b. the observed count

c. the value of the probability in the null hypothesis

d. the number of repetitions performed

c. the value of probability in the null hypothesis

the simulation done to develop the distribution we use to find our p-values assumes which hypothesis is true?

null hypothesis

when using the coin-flipping chance model, the most important repeat a simulation of the study many times is:

a. to see whether the null distribution follows a symmetric, bell-shaped curve

b. to see if the null distribution is centered around 0.5

c. to see if your coin is really fair

d. to see how much variability there is in the null distribution

d. to see how much variability there is in the null distribution

what is a standardized statistic

Z

in most statistical studies the ____ is unknown and the ___ is known

parameter; statistic

when using simulation or theory-based methods to test hypotheses about a proportion, the process of computing a p-value is

a. different id the sample is from a process instead of a finite population

b. the same if the sample is from a process instead of a finite population

c. sometimes different and sometimes the same if the sample is from a process instead of a finite population

b. the same if the sample is from a process instead of a finite population

suppose a 95% confidence interval for a population proportion is found using the 2SD or theory-based method. Which of the following will definitely be contained in that interval?

a. the population proportion

b. the sample proportion

c. the p-value

d. all of the above

b. the sample proportion

MIDTERM 2

YAYA

µ

population mean

x̄

average of the sample

population variance

σ²

sample variance

s²

sample standard deviation

s

sample standard deviation of mean

s/√n

95% confidence interval

the range within which the true population mean lies, with 95% certainty

z-score formula

(x bar -M)/(sigma/√n)

t-score formula

(x bar - M)/ (s√ n)

explanatory variable

independent variable

response variable

dependent variable

what are the two types of studies

observational and experimental

oberservational

no intervention

what are two types of observational studies

retrospective and prospective

response bias

anything in survey design that influences responses

non-response bias

people don't respond

lurking variable

is connecting x and y but x and y are not related

control group

individuals who receive no treatment

experimental group

individuals who receive treatment

hawthorne effect

people's performance is improved due to awareness of being observed

simple random smapling

every member of the population has an equal chance of being selected for the sample

systematic random sampling

fixed periodic sampling intervals

stratified random sampling

division of a population into smaller subgroups based off shared characteristics (i.e.: race or gender)

multistage smapling

draw a sample from a population using smaller and smaller groups at each stage

cluster sampling

divide population into subgroups, choose a subgroup at random, and choose everyone in the group

p pooled

x! + x2/ n1 + n2

to use the theory based approach, what two things have to be greater than 10

n(p) and n(1-p)

when given two samples and asked which has a larger margin of error, it will be the one

with the smaller sample size

when conducting a two-sided test of significance, the value of the parameter under the null hypothesis is not plausible and will not be contained in the 95% confidence interval when:

A. the p-value is less than or equal to 0.05

B. the p-value is greater than 0.05

A. the p-value is less than or equal to 0.05

FINAL

YAYAYA

how do you find the t score for the observed difference of two means

t = (y bar1 - ybar2)-(0)/((s1^2/n) + (s2^2/n))

how do you calculate the confidence interval given two means

y bar1- y bar2 ± t*(√(s1^2/n1) + (s2^2/n2)

when given two means with similar variances, you need to calculate whether or not

you can assume equal variance

how do you determine equal variance

s2^2/s1^2 and if it is less than 2.5 you assume equal variances

if you have equal variances you use

a pooled t test

in a paired design, each pair of observations always consists of measuring the same individual twice (T/F)

false

in a paired design, each pair of observations often consists of measuring the same individual twice (T/F)

true

in an independent groups design, each individual in a group is unrelated to all the other individuals in the study (T/F)

true

are the reaction times for students taken at 8 am compared to the reaction times taken for the same students at 10 pm a paired or non-paired study

paired

is the change in weight for those on the Atkins diet compared to the change in weight for those on the Paleo diet paired or unpaired

unpaired

null distributions using the MAD statistic will be

skewed right

when do you use the MAD statistic

when there are more than 2 proportions

MAD statistic

mean of the absolute value of differences

the MAD statistic will always be

positive

because the MAD statistic is always postive, the graph will be

right-skewed

the null is only rejected when the MAD statistic is

really large

how do you calculate chi-squared

(observed-expected)^2/expected

how do you calculate what is expected

total for row x total for column/ grand total

the higher the chi square value

the lower the p-value

how do you get DF for chi-square tests

(number of columns - 1)(number of rows-1)

correlation is based on a

linear relationship

r

correlation coefficient

how do you calculate r

slope x SDx/SDy

r^2

how much can actually be explained by the linear model