Discuss and plan interview content considerations

Question 1

Open questions

Answer 1

Draw out longer answers, they ask the respondent to think and reflect and allow the respondent to answer in their own words. Open questions ask, what, how and why

Question 2

Closed questions are closed because

Answer 2

They offer restricted scope for answering. They seek to elicit either one-word answers, or just a simple phrase and are suitable when a specific response is required, such as yes or no. Closed questions ask, when, where, and what and often begin with expressions like: do you, have you, can you, will they, is she? Closed questions may offer the respondent a choice from a restricted set of answers.

Question 3

Sequence of questions

Answer 3

3. 3.1 Example of a logical flow in question sequence in a farm management survey
   (1) screening question to determine whether the respondent grows a particular crop
   (2) facts about the management of the crop — area, land preparation, planting, fertiliser, weeding, harvest, etc
   (3) facts about yields
   (4) facts about disposition of crop — markets, storage, residues
   (5) opinions about specific management practices
   (6) important facts about total farming system — food preferences, cash flow, labour bottlenecks, etc

Question 4

Structure of the sequence of questions and social conventions

Answer 4

There should be some sort of structure to the sequence of topics, and social conventions of politeness should be accommodated in the ordering of more sensitive or controversial questions.

Within a questionnaire open-ended questions may be better placed later, for they usually require more time and thought

Question 5

Scope of questions

Answer 5

Generally speaking, we should ask people about issues where they have direct experience or interest (unless we are specifically interested in investigating the views, knowledge, and attitudes of people with varying experience and perspectives). Our questions should also be sensitive to the different types of knowledge that people have – for example, farmers may consider the size of an orchard more in terms of the number of trees than in terms of the land area, so a question about the area in hectares may elicit more of a guess than an objective and informed estimate.

Question 6

Scope of questions part 2

Answer 6

When asking questions we must make a judgement about the ability of the respondent to answer it.
• Does the respondent have experience and interest in this?
• Does his or her experience and interest match the nature and detail of the information being requested?
• Can we expect the respondent to remember this information?
• Can we expect the respondent to reveal this information?

Question 7

Recall of data by respondent

Answer 7

In considering the ability of respondents to remember data from the past, a useful distinction can be made between ‘single point’ and ‘continuous’ data and ‘registered’ and ‘non-registered’ events (Lipton & Moore, 1972).

1. Single point activities are those which are discreet and occur only once, for example, a fertiliser application;
2. Continuous activities, by contrast, may occur throughout the growing season, for example, the use of hired labour.
3. Registered activities refers to information which is more likely to have been noted (even if just mentally) such as payments in cash for goods or services rather than informal exchanges.

Question 8

Phrasing of a question

Answer 8

How a question is posed can have a serious impact on the way it is answered.

Question 9

Phrasing of a question and affirmation bias

Answer 9

How a question is posed can have a serious impact on the way it is answered.
If people are asked questions where they are unsure about the answer, they are often unwilling to seem ignorant and may answer incorrectly as a result. This is known as ‘affirmation bias’. In addition to ensuring that we should not ask questions to which we cannot reasonably expect the respondent to answer, the interviewer should make it clear that not being able to say is an acceptable response. Building rapport and confidence in the respondent as discussed above (see interview context) should assist.

Question 10

Leading questions

Answer 10

A leading question is one which directs the respondent. This is to be avoided in questionnaires, where it will bias results and in qualitative interviewing where the intention is for themes to arise from the data (rather than be planted by the researcher).
This is a very important issue particularly in semi-structured interviewing where the interviewer is not following a script and the interview has more in common with a conversation. In this context it is easy for the interviewer to fall into the conversational habits that in normal social interaction keep conversation moving.

Question 11

Language and ambiguity

Answer 11

Problems of ambiguity arise because of the complexities of language and the multiple meanings of words. Even without issues of translation into another language we have multiple meanings for words.
Questions relating to people’s livelihoods are often ambiguous because they do not consider what ‘having’, ‘holding’, ‘owning’, ‘borrowing’, ‘sharing’, ‘renting’ of land and other assets actually means in a given context. Too often, questions about asset holdings assume Western concepts of clear-cut, alienable property rights, concepts which may be inappropriate
Investigating these measures, even if they are not used in questions directly, gives a valuable ‘window’ on another culture, which sees the world in a different way from the researchers. Local concepts should be thoroughly understood and documented before using them in questionnaires.

Question 12

Double-barrelled questions

Answer 12

Avoid the use of double questions and double negatives. A double question might be ‘where do you buy and sell your maize?’ Double negatives can also be confusing

Question 13

Threatening questions and sensitive topics

Answer 13

It is important to remember that sensitive questions are culturally specific, as well as dependent on current events and politics. It may be rude in one society to ask direct questions about a topic when such questions present no problems in another

Question 14

Factual and attitudinal questions

Answer 14

Frequently, however, we want to find out about people’s attitudes and opinions, and these are much more subjective, changeable, and difficult to check or measure.
We can use
1. an attitude scale such as a Likert scale
2. Semantic scales are another scaling device; typically these employ a seven-point scale between bipolar descriptors
3. choice of statements reflecting different attitudes from which the respondent selects the one most closely fitting to his or her own view.
4. Where the research objectives are to understand rather than measure attitudes and beliefs, open questions within a semi-structured interview may be a more appropriate way of exploring respondents’ beliefs.

Question 15

Summary

Answer 15

Before including a question in a questionnaire or interview schedule, it is important to consider the respondent’s situation and ask yourself:
• Will the respondent understand the question?
• Will the respondent know the answer?
• Will the respondent reveal the answer?

Only include a question if the answer is ‘yes’ to all of the above.

Question 16

Focus Groups - conducting group interviews

Answer 16

Generally

1. In addition to the issues encountered when conducting single person interviews, group interviews need to consider the dynamics of group behaviour, and the willingness of people to disagree with others.
2. Only a few topics, say five or six, should be tackled in a group interview, and these should be prepared beforehand.
3. Focus group interviews may be more or less structured, use a moderator and their distinctive feature is the use of probing questions with a group of six to ten respondents.

Question 17

Advantages

Answer 17

1. speed and cost-effectiveness by comparison with individual interviews
2. the revelation of new issues as one respondent’s remark stimulates a discussion only partially controlled by the ‘moderator’ (interviewer)
3. the opportunity to weigh up opposing views and conflicting interests
4. the tendency for people to be more honest in the company of others who know the truth, and in a situation which is less individualised, eg concerning not your farm but farms in the area or village in general.

Question 18

Disadvantages

Answer 18

1. generalisable quantitative data are not produced
2. the moderator may impose his or her preconceptions on the group
3. the group dynamics may lead individuals to hide their true opinions
4. the group dynamics may mean some individuals do not express their views and are excluded.

Question 19

Role of the moderator

Answer 19

The role of the moderator in group interviews is to:
• introduce the subject and the research sponsors
• invite every member to express themselves openly
• convey the idea that he or she is here to learn from them
• probe with questions about what, when, where, which, and how (without turning the interview into an interrogation!)
• pace the discussion and control the dominant speakers
• record the event in as much detail as possible, preferably with the aid of an assistant.

Question 20

RPA’s (Rapid and Participatory Approaches

Answer 20

Rapid and participatory approaches rely heavily on group discussions and semi-structured interviews with key informants. They may also use visual techniques for:
• stimulating discussion
• reducing the extent to which outsiders control discussion
• recording the information provided.

it is important to remember that the same data collection considerations apply: how should information be collected and recorded, who should ‘observe’ it and record it, and when should it be collected and recorded are vitally important questions, and the same principles apply in answering these questions.