Ch13 - Designing the Data Collection Form for Communication Data

A recall loss error results from the fact that most people remember an event as having occurred more recently than it did. a. True b. False

F

An open-ended question is framed so as to give the respondent a clue as to how he or she should answer. a. True b. False

F

For many events and behaviors, the best reference time frame to avoid recall loss seems to be between two months to half a year. a. True b. False

F

Information that the respondent has forgotten creates a type of error called telescoping error. a. True b. False

F

Open-ended questions are generally used only at the end of the questionnaire. a. True b. False

F

Sensitive questions should always be put first in a questionnaire. a. True b. False

F

There is nothing wrong with the following question: "Do you believe McDonald's has fair prices, quality food, and excellent service?"

F

There is nothing wrong with the following question: "Do you favor or disapprove of increased military spending next year, which, as you know, can protect our interests abroad and help us fight the war on terrorism?" a. True b. False

F

There is nothing wrong with the following question: "How much do you think you would pay for a pair of sunglasses that will protect your eyes from the sun's harmful ultraviolet rays, which are known to cause blindness?" a. True b. False

F

Words like any, anybody, best, could, and ever are good words to use in a questionnaire. a. True b. False

F

A question that calls for two responses and creates confusion for the respondent is called a double-barreled question. a. True b. False

T

Data collection should never begin until you have pre-tested the questionnaire. a. True b. False

T

Filter questions are regularly used at the start of interviews or questionnaires to determine whether the respondent is actually a member of the population being studied. a. True b. False

T

In general, it is ideal to capture the needed data using as few questions as possible. a. True b. False

T

It's important to decide on a method of administration as early as possible because choice of method will influence the number and types of questions, wording and response categories, question sequencing, and so forth. a. True b. False

T

Order bias refers to the potential for responses to be affected by the sequence in which alternatives are presented. a. True b. False

T

Telescoping error is the tendency to remember an event as having occurred more recently than it did. a. True b. False

T

The fixed-alternative question doesn't permit individuals to explain their true position. a. True b. False

T

The most versatile type of question format used by researchers is the open-ended question. a. True b. False

T

The questionnaire pretest serves the same role in a questionnaire design that test marketing serves in new product development. a. True b. False

T

The split-ballot technique is the recommended procedure for dealing with response order bias. a. True b. False

T

The tendency for earlier questions to affect respondent's answers to later questions is called question order bias. a. True b. False

T

The words occasionally, sometimes, regularly, and often are considered ambiguous. a. True b. False

T

Which of the following is TRUE? a. A "filter" question can be used to determine if a questionnaire respondent is familiar with a certain topic. b. Any questionnaire response is a good response. c. Telescoping error refers to the fact that people tend to remember only events that have occurred rather recently. d. An optimal reference period used for framing questions is about one year. e. All of these are false.

a. A "filter" question can be used to determine if a questionnaire respondent is familiar with a certain topic.

Which of the following is TRUE? a. An advantage of scales is that a great deal of information can be secured from the respondent in a short period of time. b. Item nonresponse is the group of respondents who refuse to participate in the survey being con-ducted. c. Once a good set of questions has been developed, it is of little consequence as to the order they are asked on the questionnaire. d. All of these are true. e. None of these are true.

a. An advantage of scales is that a great deal of information can be secured from the respondent in a short period of time.

Which of the following is TRUE? a. Designing a questionnaire is often an iterative process, with steps or sequences of steps often being repeated. b. It is difficult, if not impossible, to state a question in such a way that it will mean exactly the same thing to every respondent. c. Gathering information by way of a questionnaire requires decisions with respect to structure and disguise and also whether it will be administered by mail, telephone, or personal interview. d. All of these are true. e. None of these are true.

a. Designing a questionnaire is often an iterative process, with steps or sequences of steps often being repeated.

"Did you vote in the last presidential election?" is an example of which of the following question types? a. Filter question b. Leading question c. Multichotomous question d. Double-barreled question e. Telescoping question

a. Filter question

Which of the following is TRUE when it comes to best practices for determining wording of questions? a. Use simple words b. Use leading questions c. Use unstated alternatives d. Use assumed consequences e. Use generalizations and estimates

a. Use simple words

The question, "How would you judge the price and quality of this product?" a. is double-barrelled. b. is leading. c. contains an implied alternative. d. contains an implied assumption. e. forces generalization on the part of the respondent.

a. is double-barrelled.

A question that cues the respondent to the desired answer is a. leading. b. double-barrelled. c. multichotomous. d. dichotomous. e. open-ended.

a. leading.

"Order bias" refers to the a. potential for responses to be affected by the sequence in which the response alternatives are presented. b. tendency of individuals to order their thoughts in a chronological manner. c. desire of individuals to live in an "ordered world". d. tendency of respondents to follow literally the instructions given by an interviewer. e. tendency of respondents to complete questions in the order they are presented by an interviewer.

a. potential for responses to be affected by the sequence in which the response alternatives are presented.

A split-ballot refers to a. the practice of using different phrasing or different orders for the alternatives on subsets of questionnaires to combat order bias. b. the practice of splitting the questionnaire responses in half for more accurate analysis. c. an administrative procedure in state elections. d. averaging in scale construction. e. None of these are correct.

a. the practice of using different phrasing or different orders for the alternatives on subsets of questionnaires to combat order bias.

In the randomized response model, which of the following questions would NOT be appropriate for pairing with this question: "Have you ever shoplifted?" a. "Were you born prior to 1960?" b. "Do you cheat on your income taxes?" c. "Is your birthday in September?" d. "Are you male or female?" e. All of these could be paired with the question.

b. "Do you cheat on your income taxes?"

Which of the following would NOT be an appropriate technique to use when asking sensitive questions? a. Hide the question in a group of more innocuous questions b. Alert the respondent by stating clearly that the subsequent question is of a sensitive nature c. Phrase the question in terms of other people and how they may feel or act d. Leave sensitive questions until the end of the questionnaire

b. Alert the respondent by stating clearly that the subsequent question is of a sensitive nature

In deciding to ask for specific information, researchers should ask which of the following questions? a. Is the question necessary? b. Both questions are correct. c. Do respondents have the necessary information? d. Neither question is correct.

b. Both questions are correct.

Which of the following is FALSE? a. When using the funnel approach to questionnaire design, the researcher will start with broad questions and progressively narrow the scope of further questions. b. Classification information is the heart of the survey and is so called because it will be classified into the appropriate categories of the dummy tables. c. Target information should be secured first and classification information last. d. In dealing with sensitive issues when designing a questionnaire, the preferred alternative is to avoid these types of questions altogether, if at all possible. e. One way of dealing with sensitive questions is to hide them in a group of more innocuous questions.

b. Classification information is the heart of the survey and is so called because it will be classified into the appropriate categories of the dummy tables.

Which of the following is FALSE? a. The order in which response alternatives appear can change the results on a survey question. b. Item nonresponse invalidates the results of a research project. c. When designing questionnaires, it is good to keep in mind that many people have difficulty with relatively simple tasks. d. A leading question is one formed so as to give the respondent a clue as to how he or she should answer. e. Just about any kind of information can be gathered using open-ended questions.

b. Item nonresponse invalidates the results of a research project.

What type of question is the following: "Do you feel that the quality of electronics manufactured in America are as good as those manufactured in Japan?" a. Double-barreled question b. Leading question c. Ambiguous question d. Sensitive question e. All of these are correct.

b. Leading question

Which of the following is NOT true about question sequencing? a. The first few questions should be simple and interesting to respondents. b. Narrow questions should be asked first and the questions should get progressively broader. c. Classification information should be asked last. d. Target information, which deal with the topic of the study, should be asked first. e. Questions on sensitive issues should be asked later.

b. Narrow questions should be asked first and the questions should get progressively broader.

Which of the following statements is INCORRECT with respect to the randomized-response model? a. The respondent answers one of several paired questions at random. b. The interviewer knows which question is being answered by the respondent. c. It is not possible to link responses to the paired questions with other responses such as demographic characteristics. d. The respondent is less likely to refuse to answer or to answer untruthfully if the randomized-response model is used. e. Answers to sensitive questions are more likely to be truthful.

b. The interviewer knows which question is being answered by the respondent.

Which of the following about observational forms is FALSE? a. When using observational forms, interviewer/observer bias may still be a problem. b. When constructing an observational form, it is best to loosely define what is to be observed so the observer is free to react to each individual situation. c. One of the most critical things in designing an observational form is to determine which aspects of the behavior are relevant. d. A paper-and-pencil observation form assessing purchasing behavior should parallel the logical sequence of the purchase act. e. All of these are true.

b. When constructing an observational form, it is best to loosely define what is to be observed so the observer is free to react to each individual situation.

Which of the following about questionnaire design is TRUE? a. If the early steps in questionnaire design are carefully followed, the questionnaire usually does not need to be revised. b. When designing questionnaires, the researcher should expect to do a good deal of iteration and looping among the steps. c. As a general rule, larger questionnaires are favored over smaller ones because they are less crowded. d. All of these are true. e. None of these are true.

b. When designing questionnaires, the researcher should expect to do a good deal of iteration and looping among the steps.

When determining content of individual questions, in general you want to capture the needed data using a. the same question multiple times, but worded differently. b. as few questions as possible. c. generalizations and estimates. d. sophisticated language, words, and phrases. e. All of these are correct.

b. as few questions as possible.

A researcher developing a questionnaire becomes aware of an interesting additional relationship that could be investigated at very little cost and effort by adding an additional item to the questionnaire. He knows that this is not vital to his present research purpose. He should a. investigate it to a limited extent. b. forget it. c. evaluate the cost. d. investigate it carefully. e. investigate it, as this is likely to reduce sampling error.

b. forget it.

The following is an example of a ____ question. "Should Macy's continue its excellent gift-wrapping program during Christmas? Yes ____ No ____" a. double-barreled question b. leading question c. sensitive question d. open-ended question e. None of these are correct.

b. leading question

The following is an example of a(n) ____. "Do you think gun control legislation is an effective way to end the senseless slaughter rampant in the inner cities? Yes ____ No ____" a. double-barreled question b. leading question c. ambiguous question d. implied alternative e. fixed-alternative response

b. leading question

The physical appearance of a questionnaire is important in securing respondents' cooperation, particularly for a. personal interviews. b. mail questionnaires. c. telephone interviews. d. All of these are correct. e. None of these are correct.

b. mail questionnaires.

If you were told that a young monk was rebuffed by his superior when he asked if he could "smoke while he prayed". A friend advised the monk to ask the question a different way: "May I pray while I smoke?" This illustrates that a. questions may be worded to always get a positive response. b. questions may be worded so that the response is biased by the question itself. c. questions should be worded to support the interests of the research sponsor. d. monks shouldn't smoke. e. None of these are correct.

b. questions may be worded so that the response is biased by the question itself.

A questionnaire using quality paper and printing techniques a. is an unnecessary expense. b. reflects the importance of the study. c. is unimportant in gaining respondent cooperation. d. is unimportant to questionnaire design. e. causes mistrust in the respondent.

b. reflects the importance of the study.

The first step in designing a questionnaire is to a. determine the type of questionnaire and method of administration. b. specify what information will be sought. c. determine the content of individual questions. d. determine the form of response to each question. e. determine the wording of each question.

b. specify what information will be sought.

Item nonresponse refers to a. deception on the part of a respondent. b. the respondent refusing to answer a question. c. the respondent incorrectly answering a question. d. deception on the part of the researcher. e. poor phrasing of a question.

b. the respondent refusing to answer a question.

Suppose that you need to learn how often per week Netflix customers watch movies online. Which of the following sets of response wording is the best way to capture this data? a. Response options that include "never, rarely, sometimes, often, frequently, and always" b. Response options that include "none, once a week, twice a week, three times weekly, and more than three times weekly" c. A response option that asks the respondent to fill in the blank with the number d. All of these are correct. e. None of these are correct.

c. A response option that asks the respondent to fill in the blank with the number

Which of the following is FALSE? a. Multiple-choice questions tend to bias results by the order in which the response alternatives are given. b. An important advantage of open-ended questions is that respondents may answer them in their own words and are not limited to a set of alternative answers. c. Because respondents are able to better clarify their answers with open-ended questions than with other types of questions, the results can be more easily coded by the researcher. d. All of these are false. e. None of these are false.

c. Because respondents are able to better clarify their answers with open-ended questions than with other types of questions, the results can be more easily coded by the researcher.

You have a need to determine if a respondent is actually a member of the population being studied. Which type of question would you use to do that? a. Open-ended question b. Leading question c. Filter question d. Dichotomous question e. Double-barreled question

c. Filter question

Which of the following is FALSE? a. One problem with all questionnaires is that a question and the way it is asked will affect the response. b. Initial pretests are best done by personal interview. c. Second pretests are not necessary if the initial pretest was done properly. d. All of these are false. e. None of these are false.

c. Second pretests are not necessary if the initial pretest was done properly.

For long reference periods, which of the following is FALSE? a. Telescoping error is relatively small. b. Recall loss is relatively large. c. Telescoping error may outweigh recall loss. d. The two effects—recall loss and telescoping error—counterbalance each other. e. Recall loss may outweigh telescoping error.

c. Telescoping error may outweigh recall loss.

What is wrong with this question? "What is your annual income?" (please check): ____ $10,000-$25,000/____ $25,000-$40,000/____ $40,000-$55,000/____ $55,000-$70,000 a. The categories are not of equal ranges. b. The alternatives are not realistic. c. The categories are not mutually exclusive. d. The categories are too broad. e. The categories are too narrow.

c. The categories are not mutually exclusive.

Which of the following is FALSE? a. There is a definite increase in claimed awareness when a respondent's memory is jogged using a recognition measure rather than an aided recall measure. b. One thing that affects a respondent's willingness to provide an answer is the amount of work involved in producing it. c. The only historically successful way of asking about a sensitive issue is to state that the behavior or attitude is not unusual before asking the specific questions of the respondent. d. In general it is better to address sensitive issues later, rather than earlier, in the survey. e. The randomized response model is sometimes used to secure sensitive information.

c. The only historically successful way of asking about a sensitive issue is to state that the behavior or attitude is not unusual before asking the specific questions of the respondent.

Which of the following is NOT true about the randomized-response model? a. The particular question asked is selected at random. b. The interviewer does not know which question the respondent is answering. c. The probability of the sensitive question asked is not known. d. The probability of the "innocuous" event occurring is known. e. It uses quantitative analysis.

c. The probability of the sensitive question asked is not known.

The question, "Do you exercise often? Seldom? Sometimes?" is a. leading. b. concrete. c. ambiguous. d. contains an implied alternative. e. contains an implied assumption.

c. ambiguous.

The question "How much was the price per gallon of gasoline when you last purchased it at a full-service service station?" should be avoided because it a. is beyond the respondent's ability or experience. b. uses a specific example to represent a general case. c. asks respondents for specifics when only generalities are likely to be remembered. d. is a double-barreled question. e. There is nothing wrong with the question.

c. asks respondents for specifics when only generalities are likely to be remembered.

The question in a restaurant survey, "What is your evaluation of the speed and courtesy of your waitperson?" is an example of a(n) a. unambiguous question. b. branching question. c. double-barreled question. d. generalization. e. redundant question.

c. double-barreled question.

The question, "Are you in favor of the free-choice policy for determining which school your child attends?" is an example of a(n) a. explicit alternative. b. double-barreled question. c. implicit assumption. d. leading question. e. classification question.

c. implicit assumption.

Branching questions are less appropriate with a. personal interviews. b. telephone interviews. c. mail questionnaires. d. mall intercepts interviews. e. All of these are correct.

c. mail questionnaires.

The most versatile type of question format to obtain answers from respondents is the a. dichotomous question. b. multiple-choice question. c. open-ended question. d. true-false question. e. multichotomous question.

c. open-ended question.

Information that the respondent has forgotten creates a type of error called a. telescoping. b. response latency. c. recall loss. d. systematic error. e. random error.

c. recall loss.

The purpose of a dummy table is to a. create a place to store data collected during the study. b. make sure that the study questionnaire is so simple to understand that even a "dummy" could respond. c. show how the results of the analysis will be presented. d. predict the outcome of the study. e. All of these are correct.

c. show how the results of the analysis will be presented.

The "funnel approach" refers to a. asking simple questions first. b. a method of data analysis. c. successively narrowing the focus of related questions. d. using a funnel as a visual aid when asking sensitive questions in a personal interview. e. successively broadening the scope of related questions.

c. successively narrowing the focus of related questions.

Which of the following is (are) useful in deciding what information will be sought in the development of a questionnaire? a. Randomized response model b. "Dummy tables" used to structure data analysis c. The hypotheses d. "Dummy tables" and the randized response model e. Randomized response model and the hypothesis

d. "Dummy tables" and the randized response model

An open-ended question can be used to a. seek factual information from a respondent. b. allow respondents to reply in their own words. c. uncover motivations and rich descriptions of feelings and attitudes. d. All of these are correct. e. a and c are correct.

d. All of these are correct.

Which of the following do NOT impact an individual's ability to remember information sought by a researcher? a. The presence or absence of stimuli that assist in remembering an event b. The importance of the event to be remembered c. The length of time since the event to be remembered occurred d. All of these impact an individual's ability to remember information. e. None of these impact an individual's ability to remember information.

d. All of these impact an individual's ability to remember information.

Which of the following is FALSE? a. Respondents tend to answer questions even when they do not possess the necessary information to give reasonable answers. b. Some questionnaire studies fail because the respondent is willing but unable to provide the information needed. c. A respondent may be more willing to provide information to a researcher if he or she is capable of articulating answers to the researcher's questions. d. Any response given by the respondent is good. e. Offering an incentive often affects the respondent's willingness to participate.

d. Any response given by the respondent is good.

Which of the following is NOT included in the recruiting message? a. Who you are b. Your request for help c. How long the survey will take d. Demographic questions e. Any incentives the respondent may receive

d. Demographic questions

Which of the following is TRUE? a. Branching questions are easy to develop for a mail questionnaire. b. Sequence bias is potentially more of a problem in telephone interviews than in mail questionnaires. c. With the exception of mail surveys, it is generally appropriate to number the questions on a questionnaire. d. The physical appearance of a questionnaire is especially influential in securing cooperation to studies done by mail. e. All of these are false.

d. The physical appearance of a questionnaire is especially influential in securing cooperation to studies done by mail.

Which of the following statements about leading questions is NOT true? a. They can result from accidents by careless researchers. b. They can be an indicator of advocacy research. c. They can represent an intentional attempt to manipulate a study's results. d. When used intentionally, they are excellent filter questions. e. They are written in such a way that they basically tell respondents what answer to provide.

d. When used intentionally, they are excellent filter questions.

The use of a "no opinion" option in fixed-alternative response questions is a. not necessary because individuals will always express an opinion. b. recommended by all leading authorities on questionnaire design. c. not recommended because it allows individuals to "take the easy way out". d. appropriate if pretesting or exploratory research reveals that 20-25% of the respondents either don't know or don't hold an opinion. e. not likely to affect the results of the survey.

d. appropriate if pretesting or exploratory research reveals that 20-25% of the respondents either don't know or don't hold an opinion.

In a recent McDonald's survey about new menu items, one question is, "How much do you like the sugar-free apple pies? Check one: dislike, dislike slightly, do not dislike or like, like slightly, like a great deal." This is an example of a(n) ____ question. a. open-ended b. double-barrelled c. dichotomous d. fixed-alternative response e. None of these are correct.

d. fixed-alternative response

"Please indicate your opinion on congressional spending and tax reform." This question a. would be a good opening question for a survey. b. uses the funnel approach. c. contains an implicit assumption. d. is double-barrelled. e. is dichotomous.

d. is double-barrelled.

The question, "Do you feel the government should be forced to stop picking our pockets with excessive taxes and return a portion of the taxes collected to taxpayers?" a. is leading. b. contains an implicit alternative. c. is double-barrelled. d. is leading and double-barrelled. e. is leading and contains an implicit alternative.

d. is leading and double-barrelled.

The best type of question to use to uncover motivations and rich descriptions of feelings and attitudes is a. leading. b. closed-ended. c. fixed-alternative. d. open-ended. e. projective.

d. open-ended.

John thought his exam was last week when it was actually three weeks ago. This is an example of a. implicit alternatives. b. funnel error. c. response latency. d. telescoping error. e. recall loss.

d. telescoping error.

Observation forms are generally easier to construct because a. they need not follow a logical pattern. b. they rely on observer judgment to determine response categories. c. response categories need not be exhaustive. d. the question and the way it is asked will not influence the respondent's actions. e. response categories need not be multichotomous in nature.

d. the question and the way it is asked will not influence the respondent's actions.

Which of the following does NOT affect a respondent's willingness to produce a response? a. The amount of work involved. b. The person's ability to articulate an answer. c. The sensitivity of the issue. d. The individual's ability to remember the event. e. All of these affect a respondent's willingness to provide a response.

e. All of these affect a respondent's willingness to provide a response.

Some of the problems with using close-ended questions include which of the following? a. None of the possible response alternatives may capture the respondent's true beliefs. b. Response categories must be truly exhaustive. c. Response categories must be mutually exclusive. d. Some respondents may not have an opinion. e. All of these are correct.

e. All of these are correct.

When determining the method of administration, it's important to remember that choice of method will influence a. the number and type of questions. b. question sequencing. c. the wording of questions. d. response categories. e. All of these are correct.

e. All of these are correct.

When utilizing observational methods of data collection, which of the following should be specified by the researcher? a. Who should be observed? b. What aspects of the purchase should be reported? c. When should the observations be made? d. Where should the observations be made? e. All of these are correct.

e. All of these are correct.

Which of the following are potential causes of item nonresponse? a. Poorly worded questions b. Questions that are misunderstood c. Sensitive questions d. Questions that are misinterpreted e. All of these are correct.

e. All of these are correct.

Which of the following is TRUE? a. In forming a question for a questionnaire, a question should be worded so as to secure an answer with only the required detail, not so as to generate additional information. b. It is wise to break one question into two when different frames of reference could be used by the respondent. c. The open-ended question is often employed as the beginning question in that it can provide insight into the respondent's frame of reference. d. When determining if the individual has the necessary information, the researcher should take into account how important the event was likely to be to the individual and how long ago it took place. e. All of these are correct.

e. All of these are correct.

Which of the following is TRUE? a. In the randomized response model, the interviewer must know which question is being answered to determine how many people answered the sensitive question in the affirmative. b. Behavior can be linked to demographic characteristics when using the randomized response model. c. Recall loss is a term used to describe the situation in which respondents, for whatever reasons, forget important events that have occurred recently. d. Generally speaking, the more a respondent has to work to provide an answer to a researcher's question, the higher the quality of the answer. e. All of these are false.

e. All of these are false.

Which of the following is TRUE? a. A double-barreled question combines two questions into one and thereby creates confusion for the respondent. b. It is extremely important that the first few questions on a questionnaire be simple, interesting, and in no way threatening to the respondents. c. Opinion questions are typically good openers for a questionnaire because people like to feel their opinion is important. d. None of these are true. e. All of these are true.

e. All of these are true.

A pretest of the questionnaire is useful for all of the following reasons EXCEPT a. to assess individual questions. b. to assess sequence of questions. c. to determine interviewer problems with questions. d. to determine if data collected are suitable for analysis. e. a pretest is useful for all of the above reasons.

e. a pretest is useful for all of the above reasons.

The Marketing Research Corporation of America has been hired to administer a series of "yes/no" questions to shoppers in a local mall. If questions on a survey have only two possible answers (such as "yes/no" questions), this type of questions are called a. open-ended questions. b. double-barreled questions and open-ended questions. c. close-ended questions. d. fixed-alternative response questions. e. closed-ended and fixed-alternative response questions.

e. closed-ended and fixed-alternative response questions.

The best way to ensure that the response categories offered in a fixed-alternative question are exhaustive is to a. provide an "other" category. b. provide a space for respondents to write in their answer, such as "e. ____." c. use at least ten categories. d. use at least fifteen categories. e. do extensive prior research before formulating the response categories.

e. do extensive prior research before formulating the response categories.

A person's ability to remember an event is NOT influenced by the a. length of time since the event. b. the event itself. c. stimulus used to facilitate recall. d. the importance attached to the event. e. the gender of the respondent.

e. the gender of the respondent.

In determining the type of questionnaire and the method of administration the researcher need not specify a. precisely what primary data are needed. b. how these data might be collected. c. what degree of structure will be used. d. what degree of disguise is warranted. e. the method of statistical analysis to be used.

e. the method of statistical analysis to be used.

One method of handling potentially embarrassing questions uses a. branching questions. b. the telescoping technique. c. fixed alternatives. d. item nonresponse. e. the randomized-response model.

e. the randomized-response model.