Deck 14: Data Collection in the Field, Nonresponse Error, and Questionnaire Screening

Question 1

Interviewer cheating may actually be encouraged by a compensation system that:

A) pays interviewers per completed interview
B) pays interviewers more than the competitive rate
C) pays interviewers by the hour
D) pays interviewers the competitive rate
E) pays interviewers above minimum wage

Answer 1

A

Question 2

As stated in your text, some believe that nonresponse is a problem because:

A) companies have reduced monetary incentives
B) distaste for survey participation, busy schedules, and desire for privacy
C) a more withdrawn public, involvement in night TV and the Internet
D) the new response rate formula adopted by the Marketing Research Society is much more conservative and yields lower rates than the former formula
E) None of the above; response rates are actually increasing dramatically as professional researchers have a better knowledge of what motivates potential respondents.

Answer 2

B

Question 3

Which of the following would NOT be considered a "sensitive" question?

A) one that asks a respondent's weight
B) one that asks a respondent's age
C) one that asks about a respondent's tax return information
D) one that asks about a respondent's social drinking behavior
E) one that asks about a respondent's choice of breakfast cereal

Answer 3

E

Question 4

Which of the following is NOT a strategy for dealing with hurdles in B2B respondent recruiting?

A) Recruiting gatekeepers as allies
B) Booking time on a prospect's calendar
C) Convincing a prospect that participation is important
D) Hiding the purpose of the survey from the gatekeeper and saving it only for the prospect
E) Explaining to the gatekeeper that the marketing research being conducted may eventually help the boss on the job

Question 5

An interviewer's accent, sex, and demeanor of the interviewer may be considered:

A) important for hiring purposes but have no effect on the respondent
B) are not to be considered in hiring interviewers due to equal employment opportunity laws
C) personal characteristics that can be a source of bias
D) personal characteristics that the researcher may not consider due to the Interviewer Code of Responsible Behavior Act
E) important only in the training process

Question 6

Which is true about nonsampling error?

A) It cannot be measured as can sampling error.
B) It can be measured as accurately as sampling error.
C) It can be measured more accurately than sampling error.
D) It cannot be measured in standard units.
E) It can be measured in standard units.

Question 7

Which of the following is a type of intentional fieldworker error?

A) falsehoods
B) fatigue
C) nonresponse
D) cheating
E) misunderstandings

Question 8

Interviewer misunderstanding would be considered:

A) intentional interviewer error
B) unintentional interviewer error
C) intentional fieldworker error
D) unintentional respondent error
E) intentional respondent error

Question 9

According to your text, one estimate shows the survey refusal rate of U.S. Consumers is almost:

A) 50%
B) 25%
C) 18%
D) 40%
E) 8%

Question 10

A variety of nonsampling error occurs during data collection. These data collection errors may be divided into:

A) fieldworker error and respondent error
B) data error in; data error out
C) intentional and fieldworker
D) unintentional and respondent error
E) directional and nondirectional nonsampling errors

Question 11

An interviewer gives the respondent very subtle cues such as saying "yes" to statements with which the interviewer agrees. This is an example of:

A) "yes-saying"
B) unintentional respondent fraud
C) leading the respondent
D) interviewer misunderstanding
E) misunderstandings

Question 12

An unintentional respondent error occurs when a respondent's interest in the survey wanes, and this is known as:

A) ADD
B) attention loss
C) attention sampling error
D) misunderstanding
E) leading behavior

Question 13

When it comes to validation, an industry standard is to randomly select ________ of respondents for call back.

A) 50 percent
B) 75 percent
C) 33.3 percent
D) 25 percent
E) 10 percent

Question 14

Which of the following is NOT true of respondent participation?

A) Participation is a crucial concern of marketing researchers.
B) The public may be weary of participating in research projects.
C) The public may feel intruded upon due to an increasing expectation of personal privacy.
D) The cost of conducting research has decreased as people have become resistant to participating in studies.
E) As respondent cooperation decreases, greater levels of resources must be committed to recruit comparable numbers of respondents.

Question 15

Which of the following is NOT an unintentional respondent error?

A) misunderstanding
B) guess
C) attention loss
D) nonresponse
E) distractions

Question 16

Which of the following best represents what should be done about nonsampling error?

A) Because it cannot be measured, greater attempts must be made to measure it.
B) Because it cannot be measured, researchers must know the sources of nonsampling error and learn how to minimize it.
C) Researchers must be aware that nonsampling error exists, but because it exists only in a theoretical sense, there are no known safeguards to minimize it.
D) Nonsampling error should be measured and controlled at all stages of the research process in order to ensure a 100 percent accurate study.
E) Nonsampling error should be measured and controlled at all stages of the research process in order to ensure a 95 percent accurate study.

Question 17

Falsehoods occur when:

A) respondents fail to tell the truth in surveys
B) fieldworkers fail to tell the truth in surveys
C) fieldworkers become fatigued
D) "e," allowable margin of error is "too low"
E) "e," allowable margin of error is "too high"

Question 18

Interviewer cheating may be controlled by:

A) hiring only interviewers who are honest
B) using one-way mirrors
C) supervision and validation
D) ensuring anonymity and confidentiality
E) None of the above; cheating cannot be controlled.

Question 19

Which of the following is NOT an example of how an interviewer might lead a respondent?

A) giving verbal cues such as "uh-huh" to responses the interviewer disagrees with
B) nodding "yes" to a question the interviewer agrees with
C) responding to a respondent by saying "I thought you would say 'yes' as over 90% of my respondents have agreed on this issue."
D) asking the question "Isn't conserving electricity a concern for you?"
E) asking the question "Is conserving electricity a concern for you?"

Question 20

Nonsampling error includes:

A) all types of nonresponse error
B) data gathering and handling error
C) data analysis
D) data interpretation error
E) all of the above

Question 21

The "third-person technique" is useful for:

A) reducing fatigue by bringing in another interviewer
B) reducing falsehoods and nonresponse
C) reducing leading questions because a third party is listening
D) reducing fieldworker cheating by having a third party witness the interviews
E) None of the above; there is no such "technique."

Question 22

Which of the following is not a "thorny" ethical aspect accompanying cell phone surveys?

A) Cell phone surveys are inherently unsafe.
B) Cell phone surveys are unevenly expensive for respondents.
C) Most cell phone surveys are not brief.
D) Cell phone surveys are difficult for the interviewers.
E) Cell phone users are disproportionately nonadults.

Question 23

Which of the following mechanisms is NOT mentioned by your text as commonly used by professional field data-collection companies in minimizing unintentional interviewer error?

A) selection
B) training
C) orientation sessions
D) monetary incentives
E) role playing

Question 24

Incentives often used to reduce nonresponse error in survey research include:

A) advance notification, assurance of insurance, and claims to be certified
B) post notification, monetary incentives, and follow-up contacts
C) advance notification, monetary incentives, and follow-up contacts
D) ex post facto notification, assurance of insurance, and follow-up contacts
E) promises not to "sell anything"

Question 25

Which of the following was NOT reported in the University of Michigan's Survey of Consumer Attitudes telephone surveys over the past two decades?

A) Prior to 1996, survey response rates were not declining.
B) Prior to 1996, survey response rates were declining at about 1% a year.
C) Since 1996, the survey response rates have been declining at about 1.5% a year.
D) The decline of the survey response rate is accelerating.
E) The decline of the survey response rate is finally decelerating.

Question 26

Validation is used to reduce the intentional respondent error of:

A) misunderstandings
B) distractions
C) falsehoods
D) guessing
E) third-person techniques

Question 27

Which of the following best describes the trend in nonresponse rates in telephone surveys conducted from 1979 to 2003?

A) Response rates have continually increased.
B) Response rates have remained unchanged primarily due to the effects of online surveys.
C) Response rates have slowly declined over this period of years.
D) Response rates increased dramatically from 1979 to about 1988, and then saw a downward fall that was just as dramatic.
E) Response rates increased dramatically from 1979 to about 1985, and then saw a downward fall that was just as dramatic.

Question 28

The response rate:

A) essentially enumerates the percentage of the total sample with which interviews were completed
B) essentially enumerates the percentage of the population with which interviews were completed
C) essentially enumerates the number of the total sample with which interviews were completed
D) is the percentage of positive responses to questions dealing with important research objectives
E) is the percentage of qualified respondents who are screened out of the sample due to the incidence rate

Question 29

Which of the following has been labeled the marketing research industry's biggest problem?

A) sampling error
B) falsehoods on the part of respondents
C) nonresponse error
D) item omissions
E) population misrepresentation

Question 30

Of the factors that determine whether or not a person will participate in a survey, which of those listed below was found by researchers to increase participation?

A) any topic dealing with sports
B) topics related to current TV shows that are also popular
C) topics of interest to the potential respondent
D) topics relating to taxation and social welfare
E) topics related to youth

Question 31

A respondent reaches a certain point in the interview, and then decides not to answer any more questions. This is known as:

A) quitting
B) nonresponse
C) item(s) omission
D) break-offs
E) failure to communicate

Question 32

If your response rate is 55 percent, your nonresponse rate is:

A) 55 percent
B) 45 percent
C) 55 percent/55 percent or 100 percent
D) 45 percent/55 percent or 82 percent
E) You cannot calculate the nonresponse rate by knowing only the response rate.

Question 33

Assuring anonymity or confidentiality is used as control for:

A) fatigue
B) leading the respondent
C) misunderstandings
D) interviewer cheating
E) falsehoods

Question 34

Which of the following methods used to reduce nonresponse error is best for reducing the number of not-at-homes, busy signals, and no answers?

A) advance notification
B) monetary incentives
C) increasing the "wrong targets"
D) callbacks
E) decreasing the "ineligibles"

Question 35

Reversal of scale endpoints may be useful for controlling:

A) sampling error
B) falsehoods
C) nonresponse
D) attention loss
E) misunderstandings

Question 36

Which of the following is true regarding the role professional field data collection companies play in terms of nonsampling error?

A) Thus far, these companies have failed to recognize nonsampling error.
B) Only a handful of these companies can be counted on to effectively control nonsampling error.
C) Although these companies give "lip service" to nonsampling error reduction, they offer little help in the control of nonsampling error.
D) Because their existence relies on providing high quality data, field data collection firms are relied on by researchers to know both the sources of and ways to control nonsampling error.
E) These firms are waiting on technological advances to help in the control of nonsampling error.

Question 37

What are the types of nonresponse errors?

A) refusals, break-offs, and hang-ups
B) hang-ups, refusals, and refusal to answer a particular question (item omission)
C) refusals, break-offs, and refusal to answer a particular question (item omission)
D) hang-ups, attention loss, and call back
E) hang-ups, attention loss, and misunderstandings

Question 38

One decision rule used in the research industry relating to how to define a "completed survey" is to define it as one in which all the:

A) primary questions have been answered
B) primary and secondary questions have been answered
C) primary, secondary, and tertiary questions have been answered
D) questions have been answered
E) classification questions have been answered

Question 39

Callback attempts are important to reduce nonresponse. Usually, about ________ callback attempts should be made to the not-at-homes, busy signals, and no answers.

A) 8 to 10
B) 7 to 10
C) 5 to 10
D) 3 to 10
E) 3 to 4

Question 40

"Prompters" may be used to:

A) interviewer error
B) facilitate proper training
C) control for attention loss or fatigue
D) direct interviewers where to find respondents
E) eliminate misunderstandings

Question 41

Very subtle "cues" from the interviewer may lead the respondent to answer the questions correctly. These "cues" should be properly taught to the interviewers during training so that the results may be guaranteed (within limits).

Question 42

If, after conducting a survey, it is determined that nonresponse is a problem, one alternative that may be considered is weighting. Essentially, weighting averages means:

A) applying weights (to accurately reflect the proportions of subgroups that represent the population) to subgroup, means to compute an overall score that adjusts for the nonresponse of that subgroup
B) applying weights to answers of subgroups until the answer is believed to accurately reflect the way the true subgroup in the population would have answered the research question
C) reproportioning subgroups in the population in a way that matches how they are proportioned in the sample, and then recalculating the survey data
D) applying a weighted value of .5 or less to all sample respondents' answers who are felt to be underrepresented in the sample data
E) applying a weighted value of .5 or less to all sample respondents' answers who are felt to be overrepresented in the sample data

Question 43

If an interviewer has an accent, it may bias the respondent.

Question 44

After the surveys have been completed:

A) data entry follows
B) there should be a preliminary screening of the questionnaires
C) they should be sent to the tabulation room
D) they should be shown to the client to verify that they were completed
E) all respondents should be recalled by a supervisor to ensure that they were actually interviewed

Question 45

Respondents can misunderstand, guess, lose attention, suffer distractions, and suffer respondent fatigue. These are all examples of unintentional respondent error and they all contribute to nonsampling error.

Question 46

The nonresponse rate is calculated by:

A) dividing the number of completed interviews by the number of eligible units in the sample
B) dividing the number of completed interviews into the number of eligible units in the sample
C) dividing the percentage of completed interviews by the percentage of eligible units in the sample
D) dividing the percentage of incomplete interviews by the incidence rate
E) none of the above

Question 47

When respondents have a tendency to respond favorably or unfavorably regardless of the question, it is known as:

A) yea- or nay-saying
B) endzoning
C) unreliable response
D) research with low variability
E) None of the above; the respondent is obviously favorable or unfavorable toward the topic being evaluated.

Question 48

One of the primary causes for interviewer cheating stems from the compensation method: payment per completed interview.

Question 49

Unreliable respondents, identified in questionnaire screening, should:

A) be resurveyed
B) be oversampled to reduce unreliability
C) be weighted so as to provide reliable estimates as to how their population subgroup would have responded to the survey question
D) not be encouraged to participate in future research and their incentive, if promised, should be withheld
E) be eliminated from the sample

Question 50

Nonsampling errors stem primarily from either fieldworker errors or respondent errors.

Question 51

Respondent cooperation refers to the general willingness of people to participate in research.

Question 52

Nonsampling error is defined as all errors in a survey except those attributable to the sample plan and the sample size.

Question 53

Supervision and validation are ways the marketing research industry controls for interviewers leading respondents.

Question 54

We can predetermine (and control)the level of sampling error, and we can predetermine the amount of nonsampling error we will have.

Question 55

When a respondent indicates "no opinion" to most questions, this problem type is known as which of the following?

A) unreliable responses
B) a middle-of-the-road pattern
C) nonresponse to a specific question
D) an incomplete questionnaire
E) a yea- or nay-saying pattern

Question 56

CATI stands for computer-assisted self-interviews where the respondent takes part in an online survey without any interviewer present.

Question 57

Which of the following organizations has provided the industry with a standard for calculating response rates?

A) American Marketing Association
B) International Marketing & Research Association (IMRA)
C) Council of American Survey Research Organizations (CASRO)
D) Marketing Research Association
E) International Council for Response Rate Standardization

Question 58

Supervision and validation are ways the marketing research industry controls for interviewer cheating.

Question 59

Assuring confidentiality is one way to minimize falsehoods on surveys.

Question 60

Falsehoods and nonresponse are two unintentional respondent errors.

Question 61

If a respondent is not at home or if their telephone is busy, it is standard practice in the research industry to immediately substitute that sample member with the next person on the sample frame.

Question 62

A good definition for a completed interview is one that has all questions answered up to the demographic questions at the end of the survey.

Question 63

A method that may reduce nonresponse error is to provide potential respondents with an advance notification, either by telephone or by letter, that they are going to be asked to take part in some research.

Question 64

The "third-person technique" may be used to reduce falsehoods.

Question 65

Reversal of scale endpoints is helpful in order to change the results to more closely match the research objectives. If the objectives are met, the client will have greater satisfaction and will likely use the services of the research firm again.

Question 66

An item omission refers to the failure to include an important question on a survey questionnaire. Therefore, item omissions may seriously threaten the achievement of the research objectives.

Question 67

Technology is dramatically changing data collection and harming the ability of marketing researchers to control its errors.

Question 68

Nonresponse has been labeled the marketing research industry's biggest problem.

Question 69

The more competent the field interviewer is through training, supervision, and personal skills, the lower the potential for interviewer error.

Question 70

An industry standard for validation is to recall 50 percent of the respondents to verify that they were, in fact, interviewed.

Question 71

CASRO, the Council of American Survey Research Organizations, has provided standards for calculating response rates.

Question 72

Nonresponse is composed of refusal and item omission; break-offs are not considered a form of nonresponse.

Question 73

A completed interview is always defined as an interview in which 100 percent of the questions have been answered and in a way that they are all supportive of the research objectives.

Question 74

The University of Michigan's Survey of Consumer Attitudes has shown that response rates are declining.

Question 75

In calculating the response rate for studies with screening questions and, therefore, a certain percentage of the sample who are "eligible," it is impossible to estimate the percentage of those members of the sample who are not called (i.e. not at homes, busy signals, etc.)who are "eligible." Therefore, the response rate formula omits the concept of eligibility, thus giving us response rates much lower than they should be.

Question 76

If a respondent is called and is "not at home," that sample element is no longer usable and callbacks should not be made. To do otherwise will distort the response rate.

Question 77

Response rate and nonresponse rates are related; if the response rate is 60 percent, then the nonresponse rate is 40 percent.

Question 78

It was long thought that response rates increased if the respondents have a high level of interest in the research topic. However, recent research has shown this is false.

Question 79

Response rate is calculated as the number of units in the sample divided by the number of completed interviews.

Question 80

Anonymity occurs when the respondent is given assurances that his or her answers will remain private; the researcher will not divulge the name of the respondent with the responses.