Ch 7 Inquizitive

Population of interest: Drivers on campus

Drivers on campus Getting a list of first-year, sophomore, junior, and senior campus drivers and e-mailing participants whose names you pulled out of a hat. Obtaining a sample by recruiting people whose names are pulled out of a hat is an example of simple random sampling.

Population of interest: Students who have taken a class with Professor A

Obtaining a list of all of Professor A's current and former students and selecting every fifteenth student for the study. Selecting every fifteenth student from the population of interest is an example of systematic sampling, which leads to an unbiased sample.

Population of interest: Democrats in Texas

Obtaining a list of all registered Texas Democrats and calling a sample of them through randomized digital dialing. Using randomized digit dialing is an example of using simple random sampling to obtain an unbiased sample.

Probability sampling techniques are a priority for frequency claims.

True Because frequency claims are claims about what happens in a population, it is important to use a sampling technique that produces a representative sample.

Convenience sampling is the most commonly used sampling technique.

True Convenience sampling is the most commonly used technique because it is quick and efficient. Even though it leads to a biased sample, many other steps can be taken to ensure validity in the study.

The external validity of an Internet-based survey sample depends on how the sample is recruited.

True Internet-based surveys that use random sampling could have excellent external validity.

Internet-based samples might consist of people who have stronger opinions than those who do not post their opinions online.

True People who rate something online might be different from people who do not post their thoughts online. They might have particularly strong positive or negative opinions about what they're rating, and they might also be more willing to share their opinions with strangers than people who do not post online.

Only probability sampling techniques lead to externally valid results.

True Probability sampling produces representative samples, which are externally valid.

Random assignment is only used in experimental designs.

True Random assignment is not used in correlational or frequency studies.

A random sample of 1,000 people is enough to generalize to a population of 100 million people.

True This sample size reduces the margin of error enough that most researchers feel comfortable generalizing to a very large population (as many as 330 million).

What kind of sample is best for external validity?

a sample in which each member of the population has an equal chance of being selected (In terms of external validity, the best sample is a representative sample (that is, one in which every member of the population has an equal chance of being selected).)

Researchers select five hospitals at random and survey all the nurses in each hospital.

cluster sampling Cluster sampling involves randomly selecting groups and then studying all the members of those groups.

Karim could randomly select five hospitals from his county and recruit all of the health care workers from those five hospitals.

cluster sampling This is an example of cluster sampling. Karim would be using hospitals as arbitrary clusters (rather than meaningful categories of health care workers), randomly selecting clusters, and then using all of the participants from the selected clusters.

Scenario 1: Jada is working on a study focused on attention span in children and notices that 11-year-old boys are underrepresented in her sample. After her 11-year-old brother participates in her study, she asks him to distribute flyers about the study to his male classmates and soccer team. Jada is using __________. Scenario 2: Dr. Saetang is conducting a study on the experience of being a racial minority on a college campus. He goes to the Asian Student Association, Black Student Union, and Hispanic Student Group on his campus to recruit Asian, Black, and Hispanic participants for his study. Dr. Saetang is using __________.

snowball sampling (Jada is finding a member of a group that is rare (for her sample) and asking that member to recruit others.) purposive sampling (Dr. Saetang is recruiting only certain groups of people (racial and ethnic minorities on campus) and sampling them in a nonrandom way.)

Identify the advantages of using Internet panels for polling.

Advantage You can track how participants' opinions change over time. (One advantage of Internet panels is that researchers can follow up with panelists repeatedly and track their attitudes over time.) You can avoid the problem of unanswered phone calls. (Research has shown that fewer than 6% of people answer polling calls (Kennedy & Hartig, 2019). Internet polls can address this problem by recruiting participants through mail and following up with them online.) Not advantage Everyone has Internet access to respond to the polls. (Not all panelists have Internet access. If they do not, the polling organization will offer free tablets and wireless access to include these panelists.) You don't have to use random sampling to get externally valid results. (Internet panels must be selected through random sampling to be externally valid.)

Identify the types of sampling as either applicable or not applicable to Dr. Jiménez's study.

Applicable snowball sampling (Dr. Jiménez is asking participants to recruit other participants for the study.) convenience sampling (By announcing the study in her classes, Dr. Jiménez is using a sample of people who are readily available to her.) Not applicable systematic sampling (Dr. Jiménez is not using a probability sampling strategy, and she is not randomly selecting numbers to count off and choose participants.) cluster sampling (Dr. Jiménez is not using a probability sampling strategy, and her study does not involve selecting clusters of participants.)

Researchers are planning an exit poll of the gubernatorial (governor) election to examine the attitudes of registered voters in DeKalb County in the state of Georgia. Which of the following methods for the exit poll would lead to results that would generalize to the population of interest?

Correct calling a randomly selected list of phone numbers of registered voters in DeKalb County (Calling a randomly selected list of phone numbers of registered voters is a way to obtain a random sample, which is representative of the population of interest.) recruiting every third voter leaving the polls in all precincts of DeKalb County (Recruiting every third voter leaving the polls is a way to obtain a random sample, which is representative of the population of interest.) Incorrect approaching young voters leaving the polls in all precincts of DeKalb County (Young voters are a sample that is not representative of the population of interest, because the population of interest includes people of many different ages. This would be a biased sample that would not generalize to the entire population.) recruiting voters from three precincts close to one another in DeKalb County (Because of their proximity, these three precincts may be similar to one another and may not represent the entire county. Because this is a biased sample, the results from this poll would not generalize to the population of interest.)

Which of the following approaches increase external validity in a study?

Correct using a representative sample (The best way to ensure external validity is to use an unbiased sample that represents the population of interest.) using a probability sampling technique (The best way to ensure external validity is to use an unbiased sample. Unbiased samples are obtained through probability sampling or random sampling.) Incorrect increasing the sample size (While increasing the sample size decreases the margin of error, it does not necessarily increase external validity.) using random assignment (Random assignment increases internal validity, not external validity.)

Which of the following research approaches enhance either internal or external validity?

Enhance Internal Validity random assignment (Random assignment increases internal validity because it ensures that the people in the treatment and comparison groups are not systematically different.) experimental design (True experimental designs enhance internal validity because they involve a comparison group that allows researchers to rule out alternative explanations.) Enhance External Validity probability sampling (Probability sampling enhances external validity because every member of the population of interest has an equal chance of being selected for the sample.) random sampling (Random sampling enhances external validity because it allows for the generalizability of the results to the entire population of interest.)

Nonprobability sampling techniques lead to larger sample sizes.

False Although some nonprobability sampling techniques like convenience sampling may seem like an easier way to obtain a sample, the type of sampling technique is unrelated to the sample size.

Combining sampling techniques increases bias.

False Combining multiple unbiased sampling techniques increases the representativeness of the sample.

A large sample is always more representative than a small sample.

False If researchers use biased techniques to collect a sample, then it is biased no matter the size.

External validity is always the highest priority in every type of claim.

False In many causal and association claims, external validity is valued less than internal validity.

Internet-based survey samples are always biased through self-selection.

False Internet-based samples do not always have problems with self-selection. Researchers can avoid this issue by using probability sampling techniques to recruit online participants.

Samples of people who have rated a specific product online are generally representative of the whole population of people who purchased the product.

False Internet-based samples may not be representative of the population of people who purchased the product. The fact that they took the time to post their thoughts online may indicate that they are different from people who did not.

Both probability and nonprobability sampling techniques allow researchers to generalize their results to the population of interest.

False Only results using probability sampling can be generalized to the population of interest.

If a finding does not generalize to other populations, it is invalid.

False There are many interesting findings that have no need to generalize to other populations. For example, a study conducted on people with Alzheimer's may only be concerned about generalizing the findings to patients with Alzheimer's.

Self-selection is a form of unbiased sampling.

False Volunteers are more likely to offer to participate if they are interested in the topic or benefits of the study, so using this technique leads to a biased sample.

Heejin wants to study the coping strategies used by partners of patients with breast cancer. She recruits participants by posting flyers about her study in the waiting rooms of several cancer treatment centers. Since she does not have enough participants, she then asks her participants to tell other partners of breast cancer patients about her study. Which of the following sampling methods does Heejin use in her study?

Method used purposive sampling (Heejin is recruiting only a particular type of participant: partners of patients with breast cancer. Because she is recruiting these participants by posting flyers, which is not a random technique, this is a nonprobability sampling method.) snowball sampling (Heejin is asking her participants to recruit acquaintances who are also partners of breast cancer patients.) Not method used systematic sampling (Heejin is not systematically sampling participants using two random numbers to count off participants.) convenience sampling (Heejin is not sampling participants who are easily accessible.)

Which of the following are reasons why it is important for frequency claims to use random sampling?

Reason Frequency claims make statements about a population. (To make claims about a population, you need to have a random sample, or a sample that is representative of the population.) External validity is a priority for frequency claims. (It is important for frequency claims to be able to generalize to a population.) Not Reason Frequency claims make stronger statements than association or causal claims. (A causal claim makes the strongest statement of all, but the strength of the claim is unrelated to why random sampling is important.) Random sampling requires fewer resources than random assignment. (Random sampling is relevant for frequency claims, but random assignment is not. The relevancy of these techniques is unrelated to the amount of resources needed.)

nonrandom sample self-selected sample

Relevant to Nonprobability Samples Nonprobability samples are obtained using nonrandom sampling methods, such as snowball or quota sampling. Self-selected samples are an example of a nonprobability sample because participants volunteer to join the study and therefore may not reflect the population of interest. Externally valid: unbiased, probability, random, representative sample Unknown external validity: biased, nonprobability, nonrandom, unrepresentative sample

externally valid sample unbiased sample representative sample

Relevant to Probability Samples Probability samples have good external validity because they can generalize to the population of interest. Probability samples are unbiased because they reflect the group of people that researchers are interested in studying. Probability samples are representative of the population of interest.

Which of the following samples would be representative of the population of American teenagers?

Representative an oversample of 1,250 U.S. teenagers (Oversampling is a probability sampling technique, so this would lead to a representative sample. With oversampling, researchers recruit an overrepresentative sample that they later adjust to match demographic proportions reflected in the population.) a cluster sample of 1,000 teenagers from around the United States (Cluster sampling is a probability sampling technique, so this would lead to a representative sample.) Not representative a purposive sample of 4,000 U.S. teenagers (Purposive sampling is a nonprobability sampling technique, so this would lead to a biased, unrepresentative sample. A large sample size does not indicate a representative sample.) a quota sample of 1,250 U.S. teenagers (Quota sampling often considers demographic categories in the population, but because it is a nonprobability sampling technique, this would lead to a biased, unrepresentative sample.) a snowball sample of 1,500 teenagers from around the United States (Snowball sampling is a nonprobability sampling technique, so this would lead to a biased, unrepresentative sample.) Probability sampling techniques involve random sampling, which gives each member of the population an equal chance of being selected for a study. This makes the sample representative of the population it is randomly selected from.

Bre is the president of a national organization of LGBTQIA rights in the United States. He wants to survey 1,000 members of his organization about their position on several political issues. He knows that transgender people make up only 5% of his organization, but he wants to make sure their views are accurately represented. He decides to randomly sample 100 transgender members and then adjust the final results so that transgender people are weighted to reflect their actual proportion of the organization. Identify the true and false statements about Bre's study.

True Bre is using oversampling. (Oversampling occurs when a researcher intentionally overrepresents one or more groups. Bre is using oversampling to make sure that he gets an accurate representation of transgender people, who make up a small percentage (5%) of the organization's members.) Bre is using a representative sample. (Because he is randomly selecting people from his population of interest, Bre is using a representative sample.) False Bre is using cluster sampling. (Bre is not randomly sampling subsets and then recruiting everyone from those randomly selected subsets, so he is not using cluster sampling.) Bre's study will have poor external validity. (Based on Bre's sampling method, the study results are likely to generalize to his population of interest.)

Dr. Jiménez is interested in how differing amounts of light affect how people perceive color. She finds participants for her study by making an announcement in the psychology classes she teaches at her university. After running the study, she asks each participant to tell their friends to sign up for her study. Identify the true and false statements about Dr. Jiménez's study.

True Dr. Jiménez has a nonprobability sample. (Dr. Jiménez is using nonprobability sampling techniques, which will lead to a biased sample.) Dr. Jiménez is testing a causal claim. (Dr. Jiménez is looking at how amounts of light affect color perception. The verb affect indicates that she is testing a causal claim.) False External validity is the highest priority for Dr. Jiménez's study. (Dr. Jiménez is looking to test a claim that does not prioritize external validity. Because of the kind of claim she's making and the kind of research design she's using, she does not need to be concerned about using a probability sample.) A larger sample would help increase generalizability for Dr. Jiménez's study. (A larger sample will not necessarily increase external validity because external validity is based on the sampling technique, not sample size.)

Dr. Lawrence is the director of Counseling Services at her university. She is planning to conduct a survey of 100 students to see how aware they are of the counseling services that the university offers. She wants the proportion of men and women in her sample to reflect the proportion at the university as a whole (55% women and 45% men). Dr. Lawrence plans to stand in the student center and ask people to participate until she has given the survey to 55 women and 45 men. Identify the true and false statements about Dr. Lawrence's study.

True Dr. Lawrence's study will have poor external validity. (Based on Dr. Lawrence's sampling method, she will not be able to generalize her results to her population of interest because she is not using a random sampling method.) Dr. Lawrence is using quota sampling. (Dr. Lawrence is using quota sampling by identifying subsets of the population and setting a target number for each subset in the sample.) False Dr. Lawrence is using a representative sample. (Dr. Lawrence's sample is not representative of the population because she is not using a probability sampling technique.) Dr. Lawrence is using stratified random sampling. (Dr. Lawrence is not using a random sampling technique. If she were using stratified random sampling, she would randomly select men and women for her study, rather than recruiting people in the student center and asking them to participate until fulfilling her target numbers.)

Thora is considering whether the reports from a sample of drivers who have reported accidents on a navigation app are accurate. Identify the true and false statements about this sample of drivers.

True The sample was obtained through self-selection. (Drivers who report accidents on a navigation app volunteer to participate, so this is a self-selected sample.) In this case, the sampling method does not affect the accuracy of the report. (While this sampling method could be problematic, it does not affect the accuracy of the report in this situation because accidents and traffic are the same for everybody driving on a particular road at a particular time.) False This group of drivers is a representative sample. (This group of drivers is not a representative sample because the sample was not obtained using random sampling.) This sample consists of drivers who likely have the same personality traits as other drivers who have not reported accidents on the app. (Drivers who report accidents on a navigation app likely have different personality traits than those who choose not to report. They might be more responsible and considerate of others, for example.)

convenience sampling purposive sampling snowball sampling

biased Sampling just those who are easiest to recruit adds bias because it is not representative of the overall population. However, in certain research designs it may not be a problem. Recruiting only a certain type of person can be biased because it is not representative of the overall population. However, it is necessary in some research designs. Having participants ask their acquaintances to participate in the study is biased because it is not representative of the overall population. However, it can be efficient.

Dr. Lin recruits participants from her psychology class and then randomly assigns them to one of three conditions. Ami randomly selects phone numbers for a telephone survey and then asks the people she calls to recruit additional participants. Zeynep posts advertisements for a study and then recruits every sixth person who contacts her.

biased sample Dr. Lin is using the biased technique of convenience sampling because she is using those who are most readily available to her. The fact that she then uses random assignment does not make her sample unbiased; it increases the internal validity of the research but does not affect the external validity. Even though Ami is initially using random sampling to randomly select phone numbers, she then uses snowball sampling by asking participants to recruit other participants, which adds bias to her sample. Even though Zeynep is using a strategy similar to systematic sampling by recruiting every sixth respondent, her sample is still made of people who are self-selecting to participate, so the sample is biased.

Scenario 1: A researcher at a nearby university wants to look at what teachers in a certain school district think about new policy changes. The researcher makes a list of all the schools in the district and uses a random number generator to select five schools from the district. Then the researcher interviews every teacher at each of those five schools. The researcher is using __________ in this study. Scenario 2: The campus safety committee has asked Professor Ibrahim to study bicycling on his campus. He trains two observers to rate the safety behaviors of cyclists at various locations around campus. He randomly selects 10 observation locations from the places where bicycles can be ridden on campus and randomly selects five 1-hour durations for each location. He has his observers make observations at each of the 10 places for each of the five durations. Dr. Ibrahim is using __________.

cluster sampling (In cluster sampling, clusters of participants (in this scenario, teachers at specific schools in the district) within a population of interest (all teachers in the school district) are randomly selected, and then all individuals in each selected cluster are included in the study.) multistage sampling (In multistage sampling, researchers randomly select clusters (in this scenario, locations) within a population of interest (the campus) and then randomly select "participants" (in this scenario, durations) from the clusters for observation.)

Select all of the sampling techniques that lead to an unbiased sample.

cluster sampling oversampling stratified random sampling systematic sampling multistage sampling All of these sampling methods are types of probability sampling techniques. Probability sampling techniques lead to an unbiased sample, which is representative of the population of interest.

Dr. Khan asks his intro psych students to fill out a survey on sleep quality and stress after class.

convenience sampling Dr. Khan is surveying those he can reach most easily—his students.

If a study's sample is not generalizable to the rest of the population, we would say that the study has __________ validity that is __________.

external (Generalizability to some larger population of interest is key for external validity.) poor (The lack of generalizability indicates that a certain type of validity is not being met.)

External validity is especially important for supporting __________ claims. The external validity, or generalizability, of a claim is based on the __________.

frequency (A frequency claim that you cannot generalize to a broader population or context is nearly meaningless.) sampling method (External validity is about how the sample was obtained, not about how large the sample is.)

Researchers select children at three different elementary schools at random by birth date.

multistage sampling In this instance of multistage sampling, the first stage is selecting clusters (in this case, schools). Then, in the second stage, researchers select participants (children) randomly within each cluster.

Karim could randomly select five hospitals from his county and then randomly select 50 health care workers from each of the selected hospitals.

multistage sampling This is an example of multistage sampling. Karim would be using hospitals as arbitrary clusters (rather than meaningful categories of health care workers), randomly selecting clusters, and then randomly selecting participants from the selected clusters.

A survey interested in comparing prisoners to non-prisoners includes prisoners as 10% of its sample, even though they only make up 2.5% of the total population.

oversampling Oversampling involves intentionally overrepresenting one or more groups in the sample and adjusting the final results. It allows researchers to make reliable claims about small parts of the overall population.

Karim is concerned that 15 physicians might not give him a precise statistical estimate, so he could recruit more physicians and then adjust the results later.

oversampling This is an example of oversampling. Since a sample size of 15 is quite small, Karim could intentionally oversample physicians to ensure that the results are accurate. He could later adjust the results for physicians to fit their actual proportion of the population.

A(n) __________ is the entire group of people that a researcher is interested in, while a(n) __________ is the smaller subset of that entire group that researchers study. A(n) __________ is conducted when a researcher investigates the entire group of people they are interested in.

population sample census

When __________ occurs, researchers target and recruit a particular group in a nonrandom way. When participants help recruit others to participate, __________ occurs. When researchers recruit participants from a group of people who are readily available to them, __________ occurs. All three of these techniques are __________ sampling techniques.

purposive sampling (Purposive sampling is useful when researchers want to study a specific group. For example, if researchers are interested in studying people who are in the process of quitting drinking, it would be helpful for them to recruit people at Alcoholics Anonymous meetings.) snowballing sampling (When researchers want to find rare types of people, they may ask participants to recruit their acquaintances.) convenience sampling (Convenience sampling is the most commonly used technique because of how easy and efficient it makes recruiting participants.) unrepresentative

Researchers conducting an online survey collect 50 men and 50 women in order to have equal gender representation.

quota sampling Quota sampling involves identifying subsets of the population and using nonrandom sampling strategies to fill a target number for each subset.

Karim could visit a local hospital and pass out his survey to health care workers that walk by until he reaches his goal of 15 physicians, 50 nurses, and 35 administrative staff.

quota sampling This is an example of quota sampling because Karim would be identifying meaningful categories of health care workers but using nonprobability sampling to select participants from each category.

When a study uses __________, each member of a population has an equal chance of being selected. __________ occurs when researchers randomly sample subsets of the population and then use all individuals within the subsets. __________ is often used to select participants to represent, in an unbiased way, subsets of the population that are not large enough to be accurately measured. All three of these techniques are __________ sampling techniques.

random sampling (A study that uses random sampling, or probability sampling, ensures that all members of a population of interest have an equal and known chance of being selected.) cluster sampling (In cluster sampling, researchers randomly select subsets, or clusters of the population, and use all members of the randomly sampled clusters. This differs from quota sampling, in which participants are selected nonrandomly from each cluster.) oversampling (After oversampling, researchers adjust the results so that the oversampled group is weighted to reflect its actual proportion of the population.) representative (Representative sampling techniques use probability sampling.)

Jalon conducts a study to investigate whether anonymity affects the reported political opinions of college students. He recruits 156 participants from psychology classes and divides them into two groups. He asks the first group for their opinions about political issues and tells them that their responses will be completely anonymous. He asks the second group for their political opinions but does not tell them whether or not their opinions will be kept anonymous. Label the components of Jalon's study. Not all items will have a match.

sampling method convenience sampling (Convenience sampling, or using subjects who are convenient, is the most common sampling technique in research.) population college students (The population is the larger group that Jalon is interested in.) sample psychology students (The sample is the portion of college students Jalon uses in his study.)

Mikhail is examining the level of workplace satisfaction for a multinational company with 10 offices around the world. He initially planned on studying every single member of the company, but due to budgetary restraints, he is only able to recruit participants from 3 offices. He randomly selects 3 out of the 10 total offices and then randomly selects 50 employees from each office to complete his survey. Label the components of Mikhail's study. Not all items will have a match.

sampling method multistage sampling (Mikhail is using multistage sampling because he selects two random samples: First he selects a random sample of offices, and then he selects a random sample of employees from each of those offices.) population every single employee (The population is the entire group of people the researcher is interested in studying.) sample employees who take part in Mikhail's study (The sample is the smaller set of people from the population of interest used as participants in the study.)

A political research center obtains a list of phone numbers for all registered voters in Texas and uses a random number generator to select 1,000 of the phone numbers to call. Out of 1,000 phone calls, 634 voters answer the phone. Researchers ask each voter who answers the phone which candidate for governor they plan to vote for in the upcoming election. Identify the components of the study. Not all items will have a match.

sampling method simple random sampling (Researchers put the phone numbers for all registered voters in Texas into a random number generator and randomly select 1,000.) population of interest all registered voters in Texas (The population that the researchers want to examine is registered voters in Texas who plan on voting in the upcoming election.) sample 634 registered voters (The sample in this political poll consists of the 634 registered voters that answer the phone.)

The student government at a college is interested in determining how important intercollegiate sports are to the students. Because all students have e-mail accounts, the student government can send a survey to all the students at the college. About 50% of the students respond. What is the most likely bias in this sample?

self-selection The sample in this study consists of people who self-select, or volunteer to participate. The student government cannot know whether the 50% of students who respond have particularly strong opinions or are just more willing to share.

Researchers choose students at random by selecting the last digit of their student IDs.

simple random sampling Each student in the population has an equal chance of being selected for the sample.

At the end of an online survey, participants are asked to tell their friends about it.

snowball sampling Some participants will likely tell other people about the survey, so the number of respondents will continually increase.

Karim could randomly select 15 physicians, 50 nurses, and 35 administrative staff for his sample of 100 participants.

stratified random sampling This is an example of stratified random sampling because Karim would be identifying meaningful categories of health care workers and randomly selecting participants from each category in the proportion that fits his population of interest.

Scenario 1: A college administrator knows that 30% of the students at her college are from out of state, and she wants to make sure that she maintains this proportion in her survey about admission practices at the college. She has a list of all the out-of-state and in-state students and randomly selects 30 students from the out-of-state list and 70 students from the in-state list. She is using __________. Scenario 2: The directors of an annual community concert want to learn the musical preferences of the audience. The directors choose 2 and 6 from a random number generator and place a survey card on every sixth seat beginning with the second seat. All the cards are returned as the audience leaves. They are using __________.

stratified random sampling (In stratified random sampling, the researcher selects particular demographic categories on purpose and then randomly selects individuals within each of those categories.) systematic sampling (This is systematic sampling because it involves choosing the numbers 2 and 6 randomly and then sampling every sixth member of the audience, starting at a random point 2.)

Select all of the sampling techniques that use meaningful categories from the population (e.g., demographics) and involve recruiting a certain number of participants from each of the categories in the population.

stratified random sampling (With stratified random sampling, researchers identify meaningful categories from the population of interest and randomly sample participants from each category based on its proportion of the overall population.) quota sampling (With quota sampling, researchers identify meaningful categories from the population of interest and set a target number for each category. Then, researchers select participants from each category in a nonrandom way.) oversampling (With oversampling, researchers identify meaningful categories from the population of interest and recruit extra participants from a certain category if they think the sample size for that category will be too small to make a precise estimate.)

Label each figure with the correct sampling technique. (stick figures)

stratified random sampling: tricolor, from research randomizer (Stratified random sampling involves identifying meaningful categories in a population and then randomly selecting participants from each category.) quota sampling: tricolor (Quota sampling involves identifying meaningful categories in a population and then selecting participants from each category using a nonprobability method (such as convenience or purposive sampling).) simple random sampling: b&w (Using a random number generator to select participants is an example of simple random sampling.) cluster sampling: house, all blue (Cluster sampling involves randomly selecting clusters from a population and then recruiting all of the individuals from each cluster.) multistage sampling: house, some blue (Multistage sampling involves randomly selecting clusters from a population and then randomly selecting some individuals from each cluster.)

A researcher approaches every fifth shopper who walks into a grocery store.

systematic sampling Systemic sampling involves randomly selecting numbers and using those numbers to choose one in every few people to be in the study. Systematic sampling is effective when simple random sampling may not be feasible.

These are the results of a BuzzFeed poll about superstitions with 29,000 participants. To whom can we likely generalize the results of this poll?

the 29,000 participants of the BuzzFeed poll Although this poll has a very large sample size, we can generalize the results only to the sample because it is an unrepresentative sample that was obtained through self-selection.

systematic sampling probability sampling stratified random sampling

unbiased Systematic sampling (randomly selecting two numbers: one as the starting point and the other to count off participants) is an excellent way to sample without bias when simple random sampling is not possible. Random sampling eliminates bias but can be very difficult to do. Stratified random sampling, or randomly selecting participants within each of a set of categories, is an unbiased way to sample from different groups of people.

Nasir selects telephone numbers from a random-digit dialer and then asks for the youngest male in the house who is at least 18. Travis randomly selects 15 major universities and then randomly selects 15 students from each of those universities.

unbiased sample This technique combines random sampling with systematic sampling, which allows for a representative sample. This is multistage sampling because it involves randomly selecting samples in two stages.