Statistics Unit 1
clinical trial
A medical experiment involving human subjects.
matched pairs
A type of experiment that compares just two treatments and combines matching with randomization. Each subject receives both treatments in a random order, or the subjects are matched in pairs as closely as possible, and one subject in each pair receives each treatment.
sample survey
A type of observational study in which only a few members of a particular group are studied. These group members are selected not because they are of special interest, but because they represent the larger group.
response variable
A variable that measures an outcome or result of a study.
double-blind
An experiment in which neither subjects nor physicians recording the symptoms know which treatment was received
How do the media often report observational studies?
As if there is a cause and effect relationship
What type of variable is the explanatory variable, "content of injection?" (For the Salk Vaccine Experiment)
Categorical. "Content of injection" can be either saline solution or vaccine, so it is categorical.
What type of variable is the response variable, "whether individual contracted polio?"
Categorical. "whether individual contracted polio" can be either yes or no, so it is categorical.
bias
Consistent, repeated deviation of the sample statistic from the population parameter in the same direction when we take many samples. When the design of a statistical study systematically favors certain outcomes.
When is a lurking variable often confounded with the explanatory variable?
Correct. Because subjects in an observational study self-select their treatment, often a lurking variable becomes confounded with the explanatory variable as part of this selection.
Institutional review boards (IRBs) are tasked with reviewing all studies involving human subjects to protect their rights and welfare. TRUE or FALSE: IRBs can never be overworked.
False They can easily become overworked and backlogged.
What is replication in an experiment?
Having more than one individual receive each treatment In order to measure variation, we must have at least two individuals receive each treatment. Repeating the entire experiment is NOT what is meant by replication in an experiment.
population
In a statistical study, this is the entire group of individuals about which we want information.
For valid inference, which type of data production does not require randomization? Survey Experiment None of the Above
None of the Above
What type of randomization takes place in a random survey?
Random selection of subjects
_______ are more valid measurements than __________.
Rates; counts Rates are more valid because they take into account the number of people sampled.
level of confidence
Says what percentage of all possible samples satisfy the margin of error.
convenience sampling
Selection of participants for a statistical study based on how easy they are to reach. Because this method of sampling may not take an entire population into account, it is often biased.
What is the key idea in defining confounding?
The effects of two variables cannot be separated
individuals
The objects described by a set of data. They may be people, but they may also be animals or things.
Why is a control necessary?
To determine a treatment's effect on the response variable Without a control and randomization, a researcher will not know if the treatment worked or if a lurking variable affected the response variable.
informed consent
Type of consent a subject must give, usually in writing, before participating in a study. The subject must be told in advance about the nature of the study and any risk of harm it may bring.
A student reads that we are 95% confident that the average score of young men on the quantitative part of the National Assessment of Educational Progress is 267.8 to 276.2 What statement correctly explains the interpretation of the statement read by the student?
We are 95% confident that the average NAEP score of all young men is between 267.8 and 276.2
interaction
a condition where the effect of one variable on the response variable changes depending on the level of another variable
statistic
a number, computed from sample data, estimating an unknown parameter
The scale that consistently says you are 3 pounds lighter than you actually are is an example of what?
bias
control or comparison (In a randomized comparative experiment)
comparing active treatments with control treatment OR comparing two or more treatments
A university is performing an experiment to see if sleep deprivation has an effect on test performance. They tell patients that they will then share each patient's personal results with the rest of the participants. This is violating the ethical policy of ___________.
confidentiality Confidentiality is being violated because personal information and results are being shared.
Which of the following involves two variables whose effects on the response variable cannot be separated? confounding interaction lurking variable
confounding
replication (In observational studies)
enough subjects to measure chance variation
sampling errors
errors caused by the act of taking a sample. They cause sample results to be different from the results of a census
True or False: The margin of error in a confidence statement includes sampling error, random sampling error, and non-sampling error.
false
True or false: Often the response variable and the explanatory variable are confounded.
false. Since confounding happens when the effects of two variables "on the response variable" cannot be separated, the response variable cannot be one of the two variables.
Institutional review boards (IRBs) are tasked with reviewing all studies involving human subjects to protect their rights and welfare. One criticism of the IRB process is that when they are overworked they may be tempted to categorize some studies as __________ when they shouldn't be.
minimal risk Placing a study in the minimal risk category speed up the process.
randomization (In observational studies)
no assignment of subjects to treatments; subjects usually self-select treatments
The types of errors possibly present in surveys are sampling errors, random sampling error, and ______________ errors.
nonsampling
parameter
number describing a characteristic of the population (usually unknown)
parameter or statistic? mean SAT of entering Freshman
parameter
An online store chooses an SRS of 100 customers from its list of all people who have bought something from them in the last year. It asks those selected how satisfied they are with the store's Web site. On selecting two SRSs of 100 customers at the same time, the two samples would give somewhat different results. This variation is a source of:
random sampling error Do we expect two different sets of people to answer a given question in the same manner? The correct answer to this question was actually discussed in Chapter 3.
By increasing the size of the sample, we can
reduce variability
dropouts
subjects who begin the experiment but do not complete it
What are two ways to adjust for nonsampling errors?
substitute the households weigh the responses
A drug manufacturing company was hopeful about a new drug for treating anxiety in patients who are unable to tolerate the side effects of current medications on the market. However, a randomized, double-blind trial showed that a sugar pill did almost as well as the new drug. The positive response of many individuals to an inactive treatment is called
the placebo effect
the conclusion of a confident statement always refers to
the population
block design
the random assignment of subjects to treatments is carried out separately within each block
Upon what is a treatment applied?
the subject
The Harris Poll organization has been existence since 1956, when it was founded by Louis Harris who famously was John Kennedy's pollster in 1960. Of late, they do much of their research through the Harris Poll Online panel. They use sampling plans that are "typically balanced based on age, gender, and region...income, education, and race can also be integrated." Source:http://www.harrisinteractive.com/vault/HI_SP_Sheet_SamplingSupportandDesign.pdf, June 25, 2011 Even though the sample can be carefully balanced, one caution in interpreting results is ______________.
there will be undercoverage bias. Because individuals without access cannot participate, there is undercoverage bias.
what is the purpose of the control
to help determine if the active ingredient really works
what is the purpose of a placebo
to reduce that bias that could result from subjects reacting to the treatment because they believe it will help them
True or False: Nonsampling error can be found even in a census.
true
True or False: Sampling errors are a source of error in sampling.
true
True or false: A placebo is basically a fake treatment.
true
True or False: Random sampling error is a type of sampling error.
true It is a subset of overall sampling error.
Using a local telephone book to select a simple random sample could introduce which type of bias?
undercoverage
to reduce bias, one needs to
use a random sample
randomization (In a randomized comparative experiment)
using random device to assign subjects to treatments
To adjust for nonsampling errors, once the data are collected, you can ______ the responses.
weigh
interviewer bias
when the interviewer (because of social position, poor training, etc.) influences the response in a systematic way
define respondent bias
when the respondent (person being surveyed) gives responses that influence the results in a systematic way
In view of escalating environmental degradation and incipient resource depletion, would you favor economic incentives for recycling of resource intensive consumer goods? Select the correct option(yes/no) from the drop-down list. Is the given question slanted toward a desired response?
yes What do phrases like "escalating environmental degradation" and "incipient resource depletion" signal when discussion recycling? An additional problem with a question like this is the number of people who won't understand it.
confidence statement
A statement that says how accurate our conclusions about the population are. The statement itself is comprised of the margin of error and the level of confidence.
control or comparison (In observational studies)
generally, active treatment compared with control OR two or more treatments compared
Fill in the blank: larger samples ________ than smaller samples
have less variability
In October 2011, the Gallup Poll asked a sample of 1005 adults, "Are you in favor of the death penalty for a person convicted of murder?" The proportion who said they were in favor was 61%. PART 1 Fill in the blank: Out of the 1005 people interviewed, the number of people who said they are in favor of the death penalty for a person convicted of murder is _____. (Give your answer as a whole number.) PART 2 Gallup says that the margin of error for this poll is plus/minus 4 percentage points. What correctly explains the phrase "margin of error plus/minus 4 percentage points?"
613 The results are based on a method that is usually (95% of the time) within 4 percentage points of the true value
institutional review board
A board that reviews all of an organization's planned studies in advance in order to protect the subjects from possible harm. The goal is to ensure that such studies are ethical.
exit poll
A poll in which voters are interviewed as they are leaving the voting place.
experiment
A study that deliberately imposes some treatment on individuals in order to observe their responses. The purpose is to study whether the treatment causes a change in the response.
Which of the following is NOT a principle of a good experiment? A. Comparing two active treatments B. Confounding a lurking variable with the explanatory variable C. Comparing active treatment with placebo D. Randomly assigning subjects to treatments
B. Confounding a lurking variable with the explanatory variable
It would be against basic data _______ principles to publish individual's results.
Ethic We can publish summary results, but must keep the individual's results private for reasons of ethics.
clinical trials
Experiments that study the effectiveness of medical treatments on actual patients
True or False: Confidentiality is a principle of ethical studies that is not always necessary in behavioral experiments.
False Confidentiality is necessary in human experiments. Many behavioral experiments rely on hiding the true purpose of the study because subjects may change their behaviors if they know they are being observed. Informed consent is not always necessary in social science experiments.
margin of error
How close the sample statistic lies to the population parameter.
Which of the following is an observational study? Simple Random Sample Experiment None of the Above
Simple Random Sample
What type of randomization takes place in a valid experiment?
Subjects are randomly assigned to treatment groups.
sample
The part of the population from which we collect information and is used to draw conclusions about the whole.
You are writing an opinion poll question about a proposed amendment to the Constitution. You can ask if people are in favor of "changing the Constitution" or "adding to the Constitution" by approving the amendment. Which of the two wordings will you use?
The second one, because it merely means that something is missing which needs to be added to the Constitution. Suggesting something might be wrong with the constitution...
How is randomization implemented in a good experiment?
Using chance assignment to allocate subjects to treatments
control
a treatment without the active ingredient (either no treatment or a placebo treatment) that is imposed on subjects
variable
any characteristic of an individual. It can take different values for different individuals.
replication (In a randomized comparative experiment)
applying each treatment to more than one subject in each treatment group
Jim is interested in testing the efficacy of a new drug on breast cancer. He follows 10 women who have received the drug and records their reactions to the drug. This is an example of a _____________.
observational study This is an observational study because Jim is only observing patients who are receiving the drug but he was not the one who imposed those treatments. This is not a carefully designed experiment which contains a control group.
define randomized comparative experiment
one that compares two or more treatments
Types of error are?
sampling, random sampling, and nonsampling errors
what does the margin of error do?
sets bounds on the size of the likely error due to random sampling.
A good sampling method has
small bias and small variability
News stories reporting poll results often include the statement "results for separates have a larger margin of error. This is because __________
subgroups have a smaller sample size
random sampling error
the deviation between the sample statistic and the population parameter caused by chance in selecting a random sample. The margin of error in a confidence statement includes only random sampling error.
estimation
the process of approximating the value of population parameters (using either point or interval estimates)
A final exam is a(n) __________ measurement for a student's understanding of a course.
valid While the exam might not cover everything, material in a final exam usually covers the important points of a course. Thus it is relevant to show the student's understanding of the material.
A researcher wants to see if there is a difference in sprinter speeds depending on the starting lane of a track. He will use a stopwatch to see how long it takes a sprinter to complete a 200m dash in seconds. What is the instrument of measurement?
A stopwatch A stopwatch is the instrument used to measure time to completion.
Robert wants to remodel his kitchen. He measures the width of the kitchen 3 times with a tape measure and gets a measurement of 10.2 ft, 10.4 ft, and 10.3 ft. The true width of his kitchen according to a professional is 11.1 ft. The total bias and random error for his first measurement of 10.2 ft is __________.
-0.9 ft Measured value= true value + bias + random error. So, 10.2 ft = 11.1ft + (bias + random error), meaning that (bias +random error) = 10.2 - 11.1 = -0.9 ft
What is the purpose of an experiment?
To determine whether treatment causes a change in the response
A website claims that you are 50 times more likely to die of a dog bite than a shark attack. TRUE OR FALSE: Dogs are much more dangerous than sharks.
FALSE Less people are exposed to sharks. This statistic is meaningless unless we know the numbers of these deaths relative to exposures.
Drinking water companies are required to post a water quality report annually for their customers to view. Your water company proudly reports they have never been out of compliance with EPA regulations for safe drinking water. TRUE or FALSE: This means that your drinking water has definitely never had high levels of arsenic in it.
FALSE The water company has a window of time to correct the problem before being considered "out of compliance".
A new synthetic oil company claims a 150% reduction in engine gunk. TRUE or FALSE: The oil company made a valid claim.
FALSE You can't take away more than 100% of anything.
TRUE or FALSE Another name for the explanatory variable in an experiment is lurking variable.
False A lurking variable is neither the explanatory variable nor the response variable. It is a variable that may influence the relationship between them that we cannot generally control.
True or False: It is unethical to have patients participate in clinical trials when the active drug (not the placebo) is not believed to help the patients taking the drug.
False Medical treatments can be tested in clinical trials only when there is reason to hope that they will help the patients who are subjects in the trials. This does not mean the active drug will harm the patients, just that it may not help them.
True or False A human resource representative is interested in workplace happiness at his company. He sends out a question to all employees and asks them to rate their happiness on a scale of 1 to 5. The variable of measurement in this study is the rating on the scale.
False Workplace happiness is the variable which is the result of the measurement.
To reduce the bias of a measurement, you should __________________.
improve the measuring instrument Bias can be reduced by improving the measuring instrument.
Which principle of experimentation is lacking in observational studies?
Observational studies usually have control/comparison and replication; they lack random assignment of subjects to treatments.
nonadherers
Subjects who participate but don't follow the experimental treatment
Which type of data production allows the subjects to self-select their treatments? Survey Experiment None of the Above
Survey. Individuals in a survey are not assigned to a treatment and can self-select their treatments.
TRUE or FALSE Results of an experiment may be biased if all individuals are not treated exactly the same except for the application of the actual treatments.
TRUE If the individuals in one treatment group receive an injection, but the individuals in the other group do not, trauma due to the injections may bias the results.
TRUE or FALSE In a matched pairs experiment with each subject receiving both treatments, to remove bias and/or confounding, the order in which a subject receives the treatments must be randomized.
TRUE The order of the treatments can influence the subject's response.
Which type of control/comparison allows a researcher to decide which active treatment works best?
Treatment 1 versus treatment 2
TRUE or FALSE: It is sometimes considered ethical to give a placebo to a control group when a standard treatment already exists.
True Placebo controls are approved even when standard treatments already exist.
block
a group of experimental subjects that are known before the experiment to be similar in some way that is expected to affect the response to the treatments
completely randomized design
a randomized comparative experiment where all subjects are allocated at random among all the treatments
Many experiments often suffer from bias due to __________. - refusal to participate - people who participate but do not follow the experimental treatment - participate dropout - All of the above
all of the above Refusals, nonadherers, and dropouts all can be a source of bias.
completely randomized
all the experimental subjects are allocated at random among all the treatments.
A city health clinic is providing free HIV testing. You are assigned a number on the day you have your blood drawn and you log into an internet website with your number for the results. This study is offering __________.
both anonymity and confidentially We always assume confidentiality is a must and there is anonymity since a participant only needs their assigned number to find out the results.
Ryan weighs himself on the scale five times and the scale says that he weighs 180, 180, 180, 180, and 180. He just went to the doctor and found his exact weight to be 176 lbs. Which of the following terms can be used to describe the measurements given by this scale?
both bias and reliable measurements The results systematically overstate Ryan's weight so the results are biased. The measurements are also reliable because they give the same weight every time (even though it is an incorrect weight).
True or false: In certain circumstances, cause and effect can be concluded even though random allocation of subjects to treatments is not done.
false
True or False: In order for an experiment to be consider ethical, it must only make sure that individuals give informed consent.
false Ethical experiments must have an institutional review board, informed consent, and subject confidentiality.
The three basic requirements of an ethical experiment are: the presence of an institutional review board, subject confidentiality, and ______________.
informed consent Ethical experiments must have an institutional review board, informed consent, and subject confidentiality.
A measurement of a property has predictive validity if
it can be used to predict success on tasks that are related to the property measured.
A variable is a valid measure of a property if
it is relevant or appropriate as a representation of that property.
The average of several repeated measurements of the same individual is
more reliable (less variable) than a single measurement.
double-blind experiment
neither the subjects nor the people who work with them know which treatment each subject is receiving.
A measurement process has random error if
repeated measurements on the same individual give different results. Ifthe random error is small, we say the measurement is reliable.
In a clinical trial, it is agreed that the interest of the ___________ must always prevail over the interests of science and society.
subject A subject's interest is more important than the results of the experiment. Thus, medical treatments must be administered only if they are thought to possibly help the subject.
True or false: A control is any treatment without the active ingredient.
true
True or false: A placebo is one type of control.
true
True or False: In an effort to improve a statistics course, we want to interview randomly chosen students about their attitudes toward the class. Because there are 800 students currently enrolled, and six instructors who teach sections of the course, a stratified sample will give the best information.
true If students were stratified according to instructor, we can ensure that each instructor's students are represented.
define bias due to non response
when an individual chosen in the sample refuses to provide answers or can't be contacted
confidentiality
when an individual's responses in a study are not released to the public; only statistical summaries of groups of subjects are made public
anonymity
when subjects' names are not collected with the data and even the director or investigators in the study do not know which individuals the observations come from
The American Psychological Association requires consent from subjects to participate in experiments except _________.
when the behavior is observed in a public space No consent form necessary if it is a behavior observed in a public place.
The Ministry of Health and the Canadian Province of Ontario wants to know whether the national healthcare system is achieving its goal in the province. The Ministry of Health conducted the Ontario health survey, which interviewed a random sample of 61,239 people who live in the province of Ontario. The survey found that 76% of males and 86% of females in the sample had visited a general practitioner at least once in the past year. Fill in the blank: Using the quick method, the estimated margin of error for conclusions having 95% confidence about the entire adult population of Ontario is _____%. (Give your answer to one decimal place.)
0.4
Match the following sources of errors with appropriate categories. 1. Sampling error: undercoverage 2. Nonsampling error: nonresponse 3. Sampling error: a convenience sample a. The subject cannot be contacted in five calls. b. The telephone directory is used as a sampling frame. c. Interviewers choose people on the street to interview.
1-b 2-a 3-c
Match the following sources of errors with appropriate categories. 1. Sampling error: a voluntary response sample 2. Nonsampling error: a response error 3. Nonsampling error: a processing error a. The subject lies about past drug use. b. A typing error is made in recording the data. c. Data are gathered by asking people to go to a Web site and answer questions online.
1-c 2-a 3-b
A 2011 Harris poll of 2362 adults found that 1110 feet for permitting abortion under some (but not all) circumstances, a decrease of 6 percentage points from 2009. Fill in the blanks: We are 95% confident that between _____% and _____% of all adults favor the US Supreme Court decision in Roe v Wade. (Round your answers to one decimal place.)
44.9 49.1
preelection poll
A sample survey that asks people how they will vote in the future. Because people often change their minds before the election, these polls are not very reliable.
census
A sample survey that attempts to include the entire population in the sample.
voluntary response sample
A sample that chooses itself by responding to a general appeal, and is therefore often biased. Examples include write-in or call-in opinion polls.
What best describes the measurement of the response variable?
A single outcome measured on each individual or subject
observational study
A study that observes individuals and measures variables of interest, yet does not involve any intervention that will influence the responses. The purpose of such a study is to describe some group or situation.
table of random digits
A table comprised of a long string of the digits 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 with the following two properties: 1. Each entry in the table is equally likely to be any of the 10 digits 0 through 9. 2. The entries are independent of each other. That is, knowledge of one part of the table gives no information about any other part.
True or False: The margin of error of 3% implies that the confidence level is 97%.
False. The confidence level of 95% tells us what percent of the sample will lead to correct confidence intervals, not the margin of error
True or False: The size of the population determines the variability of the statistics from a random sample
False. Variability of a sample statistic depends on the size of the sample and not the size of the population
To compare 3 new fertilizers, a farmer applies them to several corn fields. Each field is divided into 3 plots, and the 3 fertilizers are randomly assigned, one to each plot within each field. Harvested corn yield is compared for the 3 fertilizers. Find the following... Individual- Treatments- Explanatory Variable- Response Variable- Control- Placebo-
Individual- each corn plot Treatments- three test fertilizers Explanatory Variable- fertilizer type Response Variable- corn yield Control- no control; all fertilizers are active, there is no dummy Placebo- if we don't have a control group, there is no need for a placebo
A common form of nonresponse in telephone surveys is "ring-no-answer." That is, a call is made to an active number but no one answers. The Italian National Statistical Institute looked at nonresponse to a government survey of households in Italy during the periods January 1 to Easter and July 1 to August 31. All calls were made between 7 and 10 P.M., but 21.4% gave "ring-no-answer" in one period versus 41.5% "ring-no-answer" in the other period. Why would a high rate of nonresponse make sample results less reliable?
Nonresponse may occur more in one sector of the population than another, introducing a confounding variable. Nonresponse may make the final sample smaller, but won't necessarily make it small. One sampling period was during the winter, and the second was during the summer. What might happen in the summer more often than winter?
Here are two opinion poll questions asked in 2007 about stem cell research. Do you support or oppose embryonic stem cell research? All in all, which is more important: conducting stem cell research that might result in new medical cures, or not destroying the potential life of human embryos involved in this research? In response to the first question, 61% supported stem cell research. But only 51% of those asked the second question thought conducting stem cell research was more important. Why do you think the second wording discouraged more people from supporting stem cell research?
The phrase "not destroying the potential life of human embryos" may have brought negative images to mind, which overshadowed the phrase "might result in medical cures." Are both questions equally fair to stem cell research? If not, what phrases make one biased?
Cheshire high school announces the result of a survey- 31% of the senior class has an MP3 player. The survey was given to a random sample of 100 seniors. What is the population?
The senior class
According to a February 2008 USA Today/Gallup Poll, 43% of Americans identify themselves as baseball fans. That is low by recent standards, as an average of 49% of Americans have said they were fans of the sport since Gallup started tracking this measure in 1993. The high point came in 1998, when Sammy Sosa and Mark McGwire pursued (and ultimately surpassed) Roger Maris's single-season home run record, at which time 56% of Americans considered themselves baseball fans. The Gallup press release says: For results based on this sample, one can say with 95% confidence that the maximum error attributable to sampling and other random effects is ±5 percentage points. Which of the following sources of error do you think have not been included in this margin of error? Sampling error Voluntary response Undercoverage Nonresponse
Undercoverage Nonresponse The margin of error only covers one potential error in sampling. Two options here really refer to sources of bias, which are not included in the margin of error.
confounding
a condition where the effects of two different variables on the response variable cannot be distinguished from each other.
placebo
a dummy treatment that outwardly resembles the active ingredient; e.g. sugar pill without an active ingredient, an injection with glucose rather than a vaccine, etc.
define probability sample
a sample chosen by chance. We must know what samples are possible and what chance, or probability, each possible sample has. Some probability samples, such as stratified samples, don't allow all possible samples from the population and may not give an equal chance to all the samples they do allow.
SRS (Simple Random Sample)
a sample into which every member of the population has equal chance of being selected
non sampling errors
are errors not related to the act of selecting a sample from the population. They can be present even in a census.
When is the response variable determined?
at the end of a study
In April 2011, a Gallup Poll asked two questions about the amount one pays in federal income taxes. Here are the two questions: Do you regard the income tax which you will have to pay this year as fair? Do you consider the amount of federal income tax you have to pay as too high, about right, or too low? One of these questions drew 57% saying the amount was fair or about right; the other, only 43%. Identify the wordings that pull the respondents away from saying the amount is about right.
fair share People can have very different opinions about a "fair share."
if you want a smaller margin of error with the same confidence
increase the size of the sample
subject
individual, particularly a person, upon which treatments are applied in an experiment
Which of the following involves two variables and how the effect of one changes the effect of the other on the response variable? confounding interaction lurking variable
interaction
lurking variable
is a variable that affects the relationship between the response variable an the explanatory variable but is not included among the variables studied
explanatory variable (x) also called factor
is a variable that we think explains or causes changes in the response variable.
treatment
is any specific experimental condition applied to the subjects. If an experiment has several explanatory variables, a treatment is a combination of specific values of these variables.
If no level of confidence is given in a confident statement, then
it is assumed that the level of confidence is 95%
Which of the following involves two variables and how the effect of one affects the relationship between the other and the response variable, but is not among those variables studied? confounding interaction lurking variable
lurking variable
The ________________ sets bounds on the size of the likely error due to random sampling.
margin of error
parameter or statistic? proportion of all students who attended the last home football game
parameter
parameter or statistic? proportion of registered voters who voted in November
parameter
An important characteristic of probability samples is that the rules of _______ apply and the size of error due to sampling is measurable.
probability
parameter or statistic? mean height of a sample of NBA basketball players
statistic
parameter or statistic? mean number of pepperoni slices on a 12" pizza from a sample of a certain brand of pepperoni pizzas
statistic
parameter or statistic? proportion of people who prefer Coke over Pepsi from a sample of mall shoppers
statistic
In an effort to improve a statistics course, we want to interview randomly chosen students about their attitudes toward the class. There are 800 students currently enrolled, and six instructors who teach sections of the course. Which sampling design would give the best information?
stratified random sampling
define bias due to question wording
when questions have leading phrases, loaded words, or ambiguities that influence the response
define bias due to under-coverage and give example
when some groups in a population are left out when the sample is chosen ex: A survey of households excludes- Homeless who can't be found People who have extremely busy lives Individuals who are in hospitals, nursing homes, motels, etc.