- Self-Completion Questionnaires Chapter 11 Exam Questions

Chapter 11 - Self-completion questionnaires

Test Bank

Type: true-false

Title: Chapter 11 - Question 01

01) A self-completion questionnaire is sometimes referred to as a self-administered questionnaire.

a. True

b. False

Type: true-false

Title: Chapter 11 - Question 02

02) Self-completion questionnaires generally contain fewer closed than open questions.

a. True

b. False

Type: multiple choice question

Title: Chapter 11 - Question 03

03) Which of the following is not an advantage of a self-completion questionnaire?

Quicker to administer: Self-completion questionnaires can be sent out by post or otherwise distributed in very large quantities at the same time. A thousand questionnaires can be sent out by post in one batch, but, even with a team of interviewers, it would take a long time to conduct personal interviews with a sample of that size. However, it is important to bear in mind that the questionnaires do not all come back immediately and that they may take several weeks to be returned. Also, there is invariably a need to send out follow-up letters and/or questionnaires to those who fail to return them initially, an issue that will be returned to below.

Absence of interviewer effects: It was noted in Chapter 10 that various studies have demonstrated that characteristics of interviewers (and respondents) may affect the answers that people give. While the findings from this research are somewhat equivocal in their implications, it has been suggested that such characteristics as ethnicity, gender, and the social background of interviewers may combine to bias the answers that respondents provide. Obviously, since there is no interviewer present when a self-completion questionnaire is being completed interviewer effects are eliminated. However, this advantage probably should be regarded fairly cautiously, since few consistent patterns have emerged over the years from research to suggest what kinds of interviewer characteristics produce a bias in answers. Probably of greater importance is the tendency for respondents to be more likely to exhibit social desirability bias when an interviewer is present.

No interviewer variability: Self-completion questionnaires do not suffer from the problem of interviewers asking questions in a different order or in different ways.

Convenience for respondents: Self-completion questionnaires are more convenient for respondents, because they can complete a questionnaire when they want and at the speed that they want to go.

a. Cheaper to administer

b. More accurate responses

c. Absence of interviewer effects

d. No interviewer variability

Type: multiple response question

Title: Chapter 11 - Question 04

04) Outline two disadvantages of a self-completion questionnaires. Please select all that apply.

Cannot probe: There is no opportunity to probe respondents to elaborate an answer. Probing can be very important when open-ended questions are being asked. Interviewers are often trained to get more from respondents. However, this problem largely applies to open questions, which are not used a great deal in self-completion questionnaire research.

Cannot ask many questions that are not salient to respondents: Respondents are more likely than in interviews to be- come tired of answering questions that are not very salient to them, and that they are likely to perceive as boring. Because of the risk of a questionnaire being consigned to a waste paper bin, it is important to avoid including many non-salient questions in a self- completion questionnaire. However, this point suggests that, when a research issue is salient to the respondent, a high response rate is feasible (Altschuld and Lower 1984). This means that, when questions are salient, the self-completion questionnaire may be a good choice for researchers, especially when the much lower cost is borne in mind.

Difficulty of asking other kinds of question: In addition to the problem of asking many questions that are not salient to respondents, as previously suggested, it is also important to avoid asking more than a very small number of open questions (because respondents frequently do not want to write a lot). Questions with complex structures, such as filters, should be avoided as far as possible (because respondents often find them difficult to follow).

Questionnaire can be read as a whole: Respondents are able to read the whole questionnaire be- fore answering the first question. When this occurs, none of the questions asked is truly independent of the others. It also means that you cannot be sure that questions have been answered in the correct order. It also means that the problems of question order effects, of the kind discussed in Chapter 9, may occur.

Do not know who answers: With postal questionnaires, you can never be sure that the right person has answered the questionnaire. If a questionnaire is sent to a certain person in a house- hold, it may be that someone else in that household completes the questionnaire. It is also impossible to have any control over the intrusion of non-respondents (such as other members of a household) in the answering of questions. Similarly, if a questionnaire is sent to a manager in a firm, the task may simply be delegated to someone else. This advantage of the structured inter- view over the postal questionnaire does not apply when the former is administered by telephone, since the same problem applies. There is a feeling among some commentators that when a self-completion questionnaire is administered over the Internet (see Chapter 28 for more on this), the problem of not knowing who is replying is exacerbated because of the propensity of some Web users to assume online identities (Couper 2004).

Cannot collect additional data: With an interview, interviewers might be asked to collect snippets of information about the workplace, firm, manager, or whatever. This is not going to be possible in connection with a postal questionnaire, but, if self-completion questionnaires are handed out in an organization, it is more feasible to collect such additional data.

Difficult to ask a lot of questions: As signalled above, because of the possibility of ‘respondent fatigue’, long questionnaires are rarely feasible. They may even result in a greater tendency for questionnaires not to be answered in the first place, since they can be daunting.

Not appropriate for some kinds of respondent: Respondents whose literacy is limited or whose facility with English is restricted will not be able to answer the questionnaire, as illustrated by the example in Research in focus 8.9 of the exclusion of manual workers in a cement factory from a questionnaire survey owing to low levels of literacy. The second of these difficulties cannot be entirely overcome when interviews are being employed, but the difficulties are likely to be greater with postal questionnaires.

Greater risk of missing data: Partially answered questionnaires are more likely, be- cause of a lack of prompting or supervision, than in inter- views. It is also easier for respondents actively to decide not to answer a question when on their own than when being asked by an interviewer. For example, questions that appear boring or irrelevant to the respondent may be especially likely to be skipped. If questions are not answered, this creates a problem of missing data for the variables that are created.

Lower response rates: One of the most damaging limitations is that surveys by postal questionnaire typically result in lower response rates (see Key concept 8.5) than comparable interview- based studies. The significance of a response rate is that, unless it can be proven that those who do not participate do not differ from those who do, there is likely to be the risk of bias. In other words, if, as is likely, there are differences between participants and refusals, it is probable that the findings relating to the sample will be affected. If a response rate is low, it seems likely that the risk of bias in the findings will be greater.

a. Cannot probe

b. Cannot research sensitive areas

c. Do not know who answers

d. Less convenient for respondents

Type: multiple response question

Title: Chapter 11 - Question 05

05) Outline some steps to improve response rates to postal questionnaires. Please select all that apply.

Write a good covering letter explaining the reasons for the research, why it is important, and why the recipient has been selected; mention sponsorship if any, and provide guarantees of confidentiality. The advice provided in Tips and skills ‘Topics and issues to include in an introductory statement’ (Chapter 9) in connection with the kind of letter that might go out in advance of a respondent being asked to be interviewed can also be followed to good effect in connection with a postal questionnaire.

Postal questionnaires should always be accompanied by a stamped addressed envelope or, at the very least, return postage.

Follow up individuals who do not reply at first, possibly with two or three further mailings. The importance of reminders cannot be overstated—they do work. Our preferred and recommended approach is to send out a reminder letter to non-respondents two weeks after the initial mailing, reasserting the nature and aims of the survey and suggesting that the person should con- tact either the researcher or someone else in the research team to obtain a replacement copy of the questionnaire if the initial mailing has been mislaid or lost. Then, two weeks after that, all further non- respondents should be sent another letter along with a further copy of the questionnaire. These reminders have a demonstrable effect on the response rate. Some writers argue for further mailings of reminder letters to non-respondents. If a response rate is worryingly low, such further mailings would certainly be desirable. Some of the tactics used by Fey and Denison (2003; see Research in focus 10.1) can also be used.

Unsurprisingly, shorter questionnaires tend to achieve better response rates than longer ones. However, this is not a clear-cut principle, because it is di cult to specify when a questionnaire becomes ‘too long’. Also, the evidence suggests that the effect of the length of questionnaires on response rates cannot be separated very easily from the salience of the topic(s) of the research for respondents and from the nature of the sample. Respondents may be highly tolerant of questionnaires that contain many questions on topics that interest them.

Clear instructions and an attractive layout improve postal questionnaire response rates. Dillman (1983), as part of what he calls the Total Design Method (TDM) for postal questionnaire research, recommends lower case for questions and upper case for closed-ended answers. However, with the growing use of electronic communication and the associated rise of ‘netiquette’, upper case is increasingly associated with shouting, so that this recommendation may become less desirable.

Do not allow the questionnaire to appear unnecessarily bulky. Dillman (1983) recommends a booklet

for- mat for the questionnaire and using the photocopier to reduce the size of the questionnaire to fit the booklet format. This also gives the impression of a more professional approach.

As with structured interviewing (see Chapter 9), begin with questions that are more likely to be of interest to the respondent. This advice is linked to the issue of salience (see above) but has particular significance in the context of research that may have limited salience for the respondent.

There is some controversy about how significant for response rates it is to personalize covering letters by including the respondent’s name and address (Baumgartner and Heberlein 1984). However, one of the features of the TDM approach advocated by Dillman (1983) is that these details are supplied on covering letters and each is individually signed.

We are inclined to the view that, in general, postal questionnaires should comprise as few open questions as possible, since people are often deterred by the prospect of having to write a lot. In fact, many writers on the subject recommend that open questions are used as little as possible in self-completion questionnaires.

Providing monetary incentives can be an effective way of increasing the response rate, although it is very unlikely to be an option for most students undertaking project work or research for their dissertation. Incentives are more effective if the money comes with the questionnaire rather than if it is promised once the questionnaire has been returned. Apparently, respondents typically do not cynically take the money and dis- card the questionnaire! The evidence also suggests that quite small amounts of money have a positive impact on the response rate, but that larger amounts do not necessarily improve the response rate any further.

a. Write a strong covering letter

b. Write to respondents regularly to ensure they respond

c. Do not allow the questionnaire to appear unnecessarily long or bulky

d. Send them a gift in advance to entice them to complete the survey

Type: true-false

Title: Chapter 11 - Question 06

06) Many researchers in designing a questionnaire prefer a horizontal format because confusion can arise with a vertical format.

a. True

b. False

Type: multiple choice question

Title: Chapter 11 - Question 07

07) Which of the following is not an advantage of an email or online survey compared to a postal questionnaire?

• Low cost
• Faster response
• Attractive formats
• Mixed Administration
• Unrestricted compass
• Fewer unanswered questions
• Better response to open questions
• Better data accuracy

a. Low cost

b. Attractive format

c. More accurate response

d. Better response to open questions

Type: multiple response question

Title: Chapter 11 - Question 08

08) Which of the following are disadvantages of an online or email survey compared to a postal questionnaire? Please select all that apply.

• Low response rate
• Restricted to online populations
• Requires motivation
• Confidentiality and anonymity issues.
• Multiple replies

a. Low response rate

b. Wide sample of individuals

c. Confidentiality and anonymity issues

d. Less data accuracy

Type: true-false

Title: Chapter 11 - Question 09

09) A personal diary can be used as a method of data collection in social scientific research.

a. True

b. False

Type: multiple choice question

Title: Chapter 11 - Question 10

10) Which of the following is not a recommendation of Corti (1993) in preparing a diary study?

• provide explicit instructions for diarists;
• be clear about the time periods within which behaviour is to be recorded—for example, one working day, 24 hours, one week;
• provide a model of a completed section of a diary;
• provide checklists of ‘items, events, or behaviour’ that can jog people’s memory—but the list should not become too daunting in length or complexity;
• include fixed blocks of time or columns showing when the designated activities start and finish: for example, diaries of the kind used by Stewart (1967), which showed how managers spend their time.

a. Provide explicit instructions

b. Be clear about time periods within which behaviour is to be recorded

c. Give very prescriptive instructions as to the type of answers diarists should supply

d. Provide a model of a completed section of a diary

Type: multiple choice question

Title: Chapter 11 - Question 11

11) Which of the following is not an advantage of using a diary as a method of data collection?

• When fairly precise estimates of the frequency and/or amount of time spent in different forms of behaviour are required, the diary may provide more valid and reliable data than questionnaire data (see Research in focus 10.6).
• When information about the sequencing of different types of behaviour is required, the diary is likely to perform better than questionnaires or interviews.

The first two advantages could be used to suggest that structured observation would be just as feasible, but structured observation is probably less appropriate for producing data on behaviour that is personally sensitive, such as work-related gossip (see Research in focus 10.7). Moreover, although data on such behaviour can be collected by structured interview, it is likely that respondents will be less willing to divulge personal details. If such information were collected by questionnaire, there is a greater risk of recall and rounding problems.

a. When precise estimates of the frequency or type of behaviour are required

b. When information about the sequencing of different types of behaviour and attitude is required

c. For providing an objective assessment of one’s behaviour or attitudes

d. Producing data which is of a sensitive nature

Type: multiple response question

Title: Chapter 11 - Question 12

12) Which of the following are disadvantages of using a diary method? Please select all that apply.

• They tend to be more expensive than personal interviews;
• They suffer high levels of attrition.

a. They tend to be more expensive than personal interviews

b. They are less accurate than surveys or interviews

c. They cannot unlock as profound insights as structured observation

d. They suffer high levels of attrition

Type: true-false

Title: Chapter 11 - Question 13

13) In experience sampling, participants are prompted to reply to questions about their behaviour at particular points in time.

a. True

b. False

Type: multiple choice question

Title: Chapter 11 - Question 14

14) Which of the following is not a type of experience sampling?

1. interval contingent—where responses are provided at predetermined intervals, e.g. every hour, or at the same time each day, such as in Wagner et al.’s (2013) study of emotional labour and bus drivers;
2. event contingent—participants respond when the event takes place, such as when they experience certain moods;
3. signal contingent—participants are prompted to respond by a signalling device, such as an alarm, at randomly selected points in the day, as in Hofmans’ (2013) study mentioned above.

a. Interval contingent

b. Optimal contingent

c. Event contingent

d. Signal contingent

Type: true-false

Title: Chapter 11 - Question 15

15) Experience sampling studies may offer the researcher greater ecological validity to their work.

a. True

b. False