Once the questionnaire has been finalized, the data collection developer should begin preparing for data collection. There are a number of preparatory activities that are advised prior to the commencement of fieldwork. Below are the activities that should be followed.

7.1.1 Publicity Campaign

The purpose of the publicity campaign is to raise awareness among the population of the forthcoming data collection activity, to help them understand the significance of the activity, and to encourage them to answer all the questions honestly.

Another important task of a publicity campaign is to introduce the agency responsible for the data collection. Such an introduction will help to reassure respondents about the legality of data collection, and consequently reduce the number of respondent refusals. Publicity can be done in a number of ways:

1. Conduct an awareness campaign.
   This can be done through participation in parades, programs held during holidays, and other related activities where lots of people gather together.
2. Have a regular column in a newspaper.
   This is one way of informing the public on the upcoming data collection activity. The answers to most frequently asked questions and some mechanics of data collection can be stated in the column. Of course, the explanation should be simple enough for the general public to understand.
3. Have regular segments in the broadcast media.
   Among the different media, radio is the most effective medium in developing and under-developed countries because it can reach remote areas and those in the low-income groups. For a nationwide undertaking, the local officers of the data collection agency should coordinate with the local radio broadcasters. The cost to be incurred through this means is also low compared to the print and television media.
4. Produce calendars and posters for distribution.
   Calendars and posters can include information about the survey/census. These can be distributed in the provincial offices, towns, villages, or to the households.
5. Prepare brochures for distribution.
   Brochures containing facts about the different information to be collected and its uses can be prepared and distributed.
6. Prepare streamers and banners for display.
   Streamers and banners can alert the public to the upcoming survey/census. These should be displayed in areas where a lot of people can see them.

7.1.2 Coordination with the organizations of persons with disabilities

As discussed in the earlier chapters, it is important that persons with disabilities are consulted early in the development stage of the data collection project, as they can provide valuable information about which data should be collected. During publicity campaigns, persons with disabilities should likewise be involved. Their involvement may encourage other persons with disabilities, and the organizations they represent, to participate either as respondents or as endorsers of the census or disability survey. They are the best source of information on how to conduct a campaign in accessible format and means.

7.1.3 Coordination with local officials

It is essential that coordination takes place with the local officials. They need to be informed that the survey or census will be undertaken well in advance of the actual data collection. The local officials can assist in a number of ways, one of which is in the publicity campaign. They can also provide some assistance during enumeration such as security, transportation, accommodation, meals, guides, and translators. Gaining support from the local officials is also an effective way of gaining support from the respondents. In some instances, some respondents do not cooperate if they are not given notice from the local officials.

Coordination or courtesy calls with the local officials on the first day of data collection is also important to make them aware that the data collection in their area is starting. This will also help to reduce suspicion on the part of the residents regarding the arrival of non-resident data collectors/supervisors in the area.

7.1.4 Printing and distribution of data collection instruments

Careful planning is required for prompt receipt of survey materials (e.g. questionnaires, manuals, administrative forms, and supplies) to the different areas covered in the data collection undertaking. Depending on the size of the sample households (survey) or of the total households (census), this may involve a distribution of a large volume of forms and manuals. As such, a large space for storage and a large group of individuals to do collating, sorting, and packing jobs, will be required.

Problems with the freight can arise if the central office fails to select an agency capable of delivering the forms and manuals to the various destinations on time. A possible consequence of this is that data collection begins in some areas but is delayed in others because of the non-availability of the forms. Moreover, the distance separating some areas poses a problem in distribution whenever additional forms are requested. It often takes some time before the additional forms are received, with the worst case scenario occurring when the forms do not arrive until after the enumeration period. These problems are all associated with centralized printing of data collection materials.

On the other hand, decentralized printing of forms can pose a problem regarding the standardization of the quality of printing. Due to the fact that different printers are contracted in different areas, it is possible that the quality of the printing will vary, even if the same specifications are given to the printers. This is particularly problematic if the technology for data capture is the Intelligent Character Recognition (ICR) or Optical Mark Reader (OMR) which rely on quality paper and ink for smooth operation (refer to Chapter 9 for a detailed discussion of this kind of technology).

Another important point is to make sure that the data collection instruments are made accessible for persons with diverse disabilities.

7.1.5 Recruitment

The quality of output from the field relies heavily on the effectiveness of the interviewers. Having permanent field staff may guarantee a higher quality than recruiting a new batch of field workers each time data is collected. Obviously, staff with more training and experience will be advantaged in terms of their understanding of important concepts, their ability to deal with unusual circumstances, and so on.

Although interviewers do need to be sensitised to disability issues, they do not need medical training or backgrounds. The Philippines Disability Survey of 2001-2003 used nurses and other allied medical professionals to perform functional assessments using standardised tools highly specific for the type of disability. Generally, this level of medical sophistication is not required since disability questions should be written in language that is clear and accessible to respondents.

However, interviewers should have at least completed the full period of schooling within their country and be fluent in the main language of the country. Individual countries must decide what further level of education is required as well as what formal assessments will be carried out prior to selection.

The characteristics of the interviewers (age, sex, education, professional training, employment status, past survey experience, and so on) should be recorded on a separate database. This information can then be linked to the identification numbers of interviewers for each questionnaire completed and an analysis can be carried out of individual interviewer performance.

One common problem in recruitment is the reluctance of interviewers to be assigned to a remote area. This becomes a serious problem if no qualified personnel are found in the area. One solution is to offer incentives such as a higher travelling allowance and/or higher wages.

In determining the number of interviewers needed for data collection activity, the following information is needed:

• the number of households/respondents to be covered or interviewed;
• the duration of data collection; and
• the average output of interviewer per day

The number of supervisors will depend on the extent of supervision that they will perform. Some supervisory duties include editing questionnaires, attending to problems during enumeration (refusals, resignation of interviewers etc.) and monitoring the availability of funds. In some countries, the average ratio of interviewers to supervisors is five interviewers for every supervisor.

7.1.6 Training

Thorough training is essential to ensure that both the interviewers and supervisors have the necessary knowledge and skills to collect valid and reliable data. The purpose of training is to:

• Ensure a standardized application of the survey materials
• Clarify the rationale of the study and study protocol
• Motivate interviewers
• Provide practical suggestions
• Improve the overall quality of the data

In particular, for the disability statistics the following points are important:

• Ensure understanding of comprehensive concepts of disability
• Ensure sensitivity for persons with diverse disabilities and their families

Detailed guidelines for training can be found in the WHO document, "Household Sample Surveys in Developing and Transition Countries", which is located on the UN Statistics Division's website: http://unstats.un.org/unsd/hhsurveys/index.htm.

A summary of the key points is as follows:

1. All the training should be carried out, as much as possible, by the same team to ensure consistency.
2. If possible, a booster session is recommended towards the middle of the data-collection period. The purpose of this session is to review various aspects of data collection, with a particular emphasis on those aspects that are proving complex and difficult, or those guidelines that are not being adhered to sufficiently by interviewers.
3. The training methods should include as much role playing in interviews as possible. For this to be effective, different scripts must be prepared in advance of the training so that the different branching structures of the interview, the nature of explanations that are permitted and anticipated problems during an interview with difficult respondents can be demonstrated.
4. There should be at least one opportunity, before starting the actual data collection, for trainees to conduct an interview with a real-life respondent outside of the interviewer group. The practice interviews should be tape or video-recorded as often as possible for review and feedback discussion during training sessions.
5. Training materials should be provided to all interviewers to use as reference material.

7.2.1 Interviewing techniques

The interviewer should begin by trying to gain rapport with the respondent. With a friendly and respectful manner, the interviewer should introduce themselves by name, and as a representative of the organization, presenting his or her identification card (which should be worn at all times during the interview). The survey/census should then be introduced, with an explanation of its purposes and how the information will be used.

The interviewer should be prepared to answer questions such as:

• Where did you get my name?
• Why was I chosen for this interview?
• What kinds of questions are you going to ask?
• I'm old. I'm not disabled. Why are you including me in your survey/census?
• Why does the government spend money on a survey/census instead of providing better services to those who need them?
• I am a member of (name of disability organization). Are they supporting this survey/census?
• What services are offered in my area?

Once the introduction and some explanation about the survey/census have been made, the interview can begin. The interviewer's goal is to collect accurate information by using the questionnaire according to proven interview techniques. Since data users need to combine information collected from all interviews, the questions must be presented in a uniform manner.

During the course of the interview, a few simple rules should be followed:

1. Questions should be asked exactly as worded on the questionnaire. Research has shown that even slight word changes can alter the response obtained. If questions are paraphrased or rephrased the responses cannot be accurately combined with accurately asked questions.
2. Questions must be asked in the order they appear on the questionnaire. The sequence is planned for continuity so that early questions will not affect the answers given to later questions.
3. Every question in the questionnaire must be asked. Even if the respondent, while answering one question, answers another, goes out for a while, then appears later, the interviewer should ask the second question again, in sequence (perhaps with the apology 'I know you've already told me something about this, but this next section asks...' - which acknowledges the earlier answer, but requests the respondent's co-operation in answering it again).
4. It is important that all members of the household are identified. Some households might not volunteer the names or the existence of children or others with disabilities. Ensure that everyone in the household is listed when the household composition details are collected.
5. The interviewer should wait for the respondent to finish talking before starting to record their response. Failure to listen carefully can offend the respondent and result in errors.
6. The interviewer should not interrupt the respondent, even if he or she hesitates or is quiet for a while. Sometimes people initially answer, "I don't really know" when in fact they are thinking about their answer.
7. Answers should not be anticipated. No assumptions should be made about what the respondent is going to say. The respondent should be allowed to finish their sentence.
8. If the answer is dubious and the interviewer believes the respondent does not fully understand the question, the question should be repeated as written. If the interviewer is still doubtful, probing may be necessary (see below), but the interviewer should be careful not to antagonise the respondent by questioning their judgment.
9. Questions should be asked in a positive or neutral tone. Interviewers may feel uncomfortable about a question but should refrain from voicing this. Apologetic refrains (e.g. "you might not want to answer this question, but." or "this question probably won't make much sense to you."), will only disrupt the flow of the interview and may affect how the respondent answers the question.
10. The interviewer should maintain a professional attitude and be upbeat even if a respondent is difficult or uncooperative. The interviewer should not get upset by the respondent's behaviour and be courteous throughout the interview.
11. At the end of each interview, the respondent should be thanked for their time and cooperation.
12. Ideally, the interview should be completed in one visit. With some respondents this may not be possible, and a second or even a third visit may be necessary.

Please refer to Appendix 3 for an example of interview guidelines developed for pilot studies in 2005 by WHO and ESCAP.

7.2.2 Interviewing persons with disabilities - tips and traps

Even well trained interviewers benefit from training specific to the problems and issues involved in interviewing persons with disabilities. Some tips and potential traps are listed below, grouped under common sense rules directed to the interviewer.

1. Treat persons with disabilities as you would treat anyone else

Persons with disabilities have the same needs, desires, aspirations and goals as anyone else. They differ in how they can satisfy their needs and desires, or reach their aspiration and goals.

It is a profound mistake to assume that disability changes everything about a person, or that their physical or mental differences would mean they will never understand you, or you understand them. As we know from the ICF, disability is a universal human phenomenon; limitations in functioning have been proven to be part of being human.

Persons with disabilities usually refer to daily activities in the same way others do. Do not hesitate to use the words 'look' and 'see' with people who are blind or have visually impairment, 'listen' and 'hear' with people who are deaf or have hearing impairment, or 'walk' and 'run' with people with mobility impairments.

It is also important to ask persons with disabilities on what communication needs and preference they have and accommodate the needs to the best possible extent.

2. Be respectful of persons with disabilities

We have already mentioned the need to be sensitive to language, since in most cultures, the common perceptions of persons with disabilities are distorted by stereotypes and misunderstanding.

At the same time, do not patronise persons with disabilities, or tell them that you admire their courage and strength. Establish trust and treat persons with disabilities as equal.

Never speak to a person with a disability in a tone of voice that suggests you think they are child-like or of limited intelligence. Use a normal tone of voice.

If the respondent is accompanied by an attendant or a sign language interpreter, speak directly to the person with the disability.

If a sign language interpreter is present, speak at a slower pace so that the interpreter does not miss what you say.

Fatigue may be a problem for people with head injuries or other neurological conditions. When fatigued, they may become angry or frustrated. Suggest a break or at the beginning of the interview, tell the individual to let you know when s/he would like to take a break.

3. Be aware of, and accommodate communication differences

Some impairments affect how people speak or communicate. A good interviewer will recognise these difficulties and, without being patronising, seek ways to increase communication.

Some people with communication disabilities have learned to compensate for them: persons who are deaf may be able to lip-read and converse orally.

A person with a speech impairment may have distinct speech patterns and what may appear at first to be uncontrolled can be understood after a few moments of listening. Be patient, but ask the person to repeat what s/he is saying if you do not understand. People who have unclear speech are used to people asking them to repeat things, so be honest about clarifying a response.

4. Accommodate the deaf respondent

When interviewing a deaf person it is important to have his or her attention first before speaking. A deaf person may need a tap on the shoulder, a wave of the hand, or some other visual sign to get their attention.

Speak clearly, not loudly, to a person who is hearing impaired; face them and make sure your face is well lit and not obscured by your hand or facial hair. The person will ask you to speak louder or slower if required. Do not exaggerate or overemphasize words. This distorts lip movements making lip-reading difficult.

Since some lip movements are difficult to lip-read, try to rephrase the question if the person who is lip-reading does not understand it after a couple of repetitions. It is also helpful to rephrase the question if the individual with a hearing impairment has difficulty in understanding you. Body language and facial expression are important factors in successful communication.

It is a mistake to think that people who are deaf are also non-verbal. People who are deaf have normal vocal organs but may choose not to speak. Also, not all persons who have hearing impairments can lip-read.

5. Accommodate the blind respondent

Announce your arrival and exit to a person who is blind or has a visual impairment. Identify yourself and let the person know that you are addressing them by using their name or touching their arm.

Never distract a guide dog. The dog is working and is responsible for the safety of the person who is blind.

Do not assume that all people who are blind can read Braille. If Braille materials are used, it is best to ask beforehand if the person can read them.

6. Accommodate the individual with intellectual disabilities

Persons with intellectual disabilities may need more time to understand the question. Do not rush. Make sure the person clearly understands what is being asked.

Persons with intellectual disabilities should not be treated like children; they should be treated like any other person of the same age.

Sometimes persons with intellectual disabilities may be anxious to please and say what they think you want to hear. Bear this in mind when conducting the interview.

7.2.3 Checking for completed questionnaires

After completing the interview, the interviewer should always check if all the questions are asked and if the answers are consistent. It is important to check the questionnaires before leaving the household so errors can be immediately rectified. It is clearly more difficult to correct an error after the interviewer has left the household as that involves another explanation to the respondent, as well as disruption to the interviewer's schedule and additional cost.

7.2.4 Responsibilities of the field supervisor

In any data gathering activity, supervisors play a major role in attaining high quality statistics. It is the duty and responsibility of the supervisor to:

1. Accompany interviewers, especially those new in the area of interview, in the first few days of data collection to assist them in following the proper procedures. The supervisor should observe how they conduct the interview and provide suggestions for improvement. It is important, however, for the supervisor not to intervene and correct the interviewer while the interview is being conducted, unless the error would have a major effect on the response of the respondents. In this case, the supervisor should ensure that the process of correcting the interviewer does not irritate the respondent nor reduce the credibility or morale of the interviewer.
2. Conduct spot-checking. This is the process of checking if the interviewers are actually in the area visiting the households and conducting the interview. Despite the fact that interviewers have gone through rigid screening and training procedures, it is always possible that some may be tempted not to visit all the households they are assigned to.
3. Re-interview a sub-sample of households. Supervisors should always re-interview some households to determine if the interviewers asked the correct questions to the respondents. The supervisors do not have to ask all the questions in the questionnaire, and the data collection developer may design a specific form intended for this activity. The contents normally consist of important data items in the questionnaire, such as age, sex, and few questions on disability.
   
   It is important, however, that supervisors provide clear explanation to the respondent why a re-interview is needed. It is crucial to emphasize to the respondent that they are being re-interviewed, not because the supervisor doubts the veracity of their answers, but merely to confirm some of the answers they provided. The supervisor should not tell the respondent that they are conducting a re-interview in order to check that the original interviewer was doing their job correctly, as this may call into question the interviewer's credibility.
4. Review the completed questionnaires. Supervisors should review the work of the interviewers in order to correct errors at the earliest stage possible. Early detection of errors saves incorrect information being processed, and enables the supervisor to correct the interviewer's mistake so it is not repeated in future interviews.
5. Attend to other problems during data collection such as refusals, lack of forms, supplies and materials, lost questionnaires, and replacement of staff.
6. Hold regular meetings with interviewers. Such meetings are vital as they provide a venue where problems and solutions can be discussed. They also provide an opportunity for the supervisor to discuss their observations made whilst conducting the spot-checks, re-interviews and reviews of completed questionnaires.

7.2.5 Interviewing in institutions

Interviewing people residing in institutions presents additional challenges that need to be planned for in advance.

Institution staff members (administrators or other contacts, such as head nurses and ward clerks) are busy professionals. Scheduling appointments for proxy interviews at a mutually convenient time will require flexibility. Interviewers must be prepared for the possibility that appointments may change at short notice.

When interviewing in institutions, the interviewer must be prepared to adapt the interviewing techniques to meet the requirements of a different situation. If the interviewer senses that the respondent is becoming upset, or is incapable of completing the interview, it is best to terminate the interview and discuss with the administrator or other contact person how best to continue.