The field work
By Neil Hawkins
The field work for the FALUP study took place within the general context of FAFO activities in the Middle East. An important aspect of those activities is institution building and training. During the FALCOT 92 survey, FAFO had created a core group of local staff, which served as supervisors. After the establishment of the Palestinian Bureau of Statistics, it was decided that the FAFO staff should be transferred to the new institution, in order to strengthen the capacity to carry out surveys. Therefore, FAFO decided that it would be an opportunity for the local staff to design and run as many areas of the survey work as possible. The aim was that next time they could cover all aspects of survey management and organisation themselves. In the event they handled nearly all aspects of this survey with a minimum of supervision which, was extremely impressive.
The organisation and implementation of the field work for this particular survey greatly benefited from the experiences gained by FAFO in the FAFO Living conditions Survey in 1992 (FALCOT 92). This was particularly useful as there was considerable pressure to implement the field work as soon as possible. The situation was further complicated by the closure of the Occupied Territories by the Israeli authorities which led to some staff having to "smuggle" themselves in to work in particular areas. The signing of the peace accord in Washington led to high expectations as well as wariness over anything new. Field workers often faced respondents who thought the survey was intended to resettle them or demolish their shelter.
Nevertheless the team that worked on this survey was experienced and well trained, and managed to overcome such problems. It was the hard work and dedication of the local staff that ensured the success of this survey, which took place in very difficult circumstances.
Despite having contacts with experienced data collectors from previous surveys, FAFO decided to advertise openly for the required data collectors, as FAFO did not want to be seen to be making appointments without giving all a chance to apply.
Adverts were placed in the local press and a short-list was made of all applicants after they had been screened and graded by both the Office Coordinators, and the in-field coordinator. Interviews were held in Jerusalem, Gaza and Nablus and out of the seventy-five applicants, six data collectors were chosen for the West Bank and twelve for Gaza. It must be remembered that the West Bank sample was half the size of Gaza. The majority were female and nearly all had worked with FAFO in the previous survey.
FAFO supervisors were involved both in the screening of the candidates as well as setting up and conducting the interviews.
It was decided that the most efficient way to conduct the training was to have a number of parallel courses running at the same time and led by the supervisors. The alternative was to run them consecutively (travel restrictions prevented the possibility of holding only one course for all areas) and taken by FAFO experts. This would have been time consuming and impractical. Therefore the supervisors were invited to design a course for the data collectors as well as go through the proposed questionnaire. The latter task was performed with a FAFO expert.
Meetings had already been held where the questionnaire was openly discussed in detail and the many comments of the supervisors were registered. This is a vital stage for any survey, as not only do the supervisors criticise the questionnaire based on their experience but it also is an important factor in motivation, since it is they who will have to defend the questionnaire to the data collectors. Their comments are based on field experience and contribute to preventing misunderstandings and confusion in the field work.
The questionnaire was repeatedly tested out in the field by the supervisors at this stage in order to check for any mistakes and irregularities.
Together the supervisors and in-field coordinators developed a training course. Papers were represented on the theme of training, covering the aims and various methods, and subsequently discussed in detail. Training styles and techniques were discussed and practised and a course schedule and plan was put together.
In addition to this supervisors produced a small booklet for the data collectors which gave an overview of the project, the role of the data collector, confidence building, dealing with refusals, being objective, avoiding giving the respondent expectations of help as well as interview techniques.
It was decided to use small discussion groups in the training course as this gave more opportunity for data collectors to have their say and was a more efficient use of time. The classroom section of the course would last three days. Then they would spend two days testing the questionnaire in the field to give the data collectors practice.
This stage instilled an effective feeling of teamwork as well as pride in their work, and as a result of their extensive involvement at all stages, the supervisors began to regard it more as their survey.
The training of the data collectors went smoothly and the supervisors were easily able to establish their authority through the fact that they were running the course.
The course used both the booklet and the questionnaire as a starting point but it also used role playing techniques to make the data collectors used to the various problems they would meet in the field.
Data entry operators were also trained along with the data collectors, as it was important that they understand the context of what they were doing. They also had a training course of two days given by both a FAFO expert and one of the administrative assistants. A booklet outlining how to organise and save their files, etc., was also prepared for them.
By the end of the training course FAFO were confident that the teams were ready for the field work stage.
Preparations for the survey differed between Gaza and the West Bank. In Gaza we used the same cells as the FALCOT 92 survey but seeing as we gave a commitment to the households we visited that no one would be able to use this survey to make a revisit (to avoid people worrying about having visits by the tax men), we could not go back to them. Instead we chose their neighbours as this meant we could use the basic maps again, which would save a great deal of time.
In Gaza the maps were therefore re-done based on the old ones but this was not without problems. Seeing as it was not desirable to identify houses by the names of the inhabitants, the only other way of identifying houses was by physical characteristics as there is no numbering system. Naturally, in the space of a year changes take place such as a new coat of paint, an extension or even building into the alley with the result of closing it. Nevertheless these were the exceptions and most houses were successfully identified first time round.
In the West Bank the sample population in this survey was refugee camps whereas in FALCOT 92 they were a small proportion of the total population of the West Bank sample. Therefore, FAFO had to select a new sample. Maps were obtained of the camps with the help of UNRWA and they were divided into identifiable localities (based on a road or path). A starting point was taken at random in each of the districts and marked on a map. The supervisors then went back into the field with the data collectors, found the starting point and implemented a set walking pattern to identify the selected houses. Data collectors were distributed to these houses as they walked. This method was decided upon in order to save time as the mapping had to be done prior to the collection of data, otherwise there would have been considerable delay.
For statistical aspects of the sampling, see appendix 1.
The supervisors decided that it would be better to hire the data collectors on a monthly basis instead of paying them per questionnaire. The arguments revolved around control and incentive. Being paid per questionnaire would, it is argued, lead to the data collector finishing more questionnaires, whereas if they were paid by the month there would be no incentive to finish or fill in a questionnaire well. The problem with that system is that if you need to have meetings to review questionnaires, the data collector sees that he is not getting paid for it, so he has no incentive to attend. Furthermore the system could also encourage them to work too quickly. The supervisors felt that a mixture would be best.
It was decided to pay a set salary for a month with a bonus of 18% of the total salary upon completion, depending on the quality of their work and the consistency of keeping the required records plus their attendance of meetings. This system worked very well and fostered team-work as people did not feel that a colleague was trying to do more questionnaires than anyone else, but it still left room for an incentive for good work as well as underlining the supervisors' authority.
In Gaza there were three teams each with a supervisor as well as one trouble shooter supervisor. In addition there was a sampler who played a crucial role in ensuring the accuracy of the maps. The office was headed by the Gaza coordinator and helped by her assistant who oversaw the two data enterers.
The data enterers were trained to run cleaning passes on their questionnaires and the print out was given to the respective office coordinator. They would then return the inconsistent questionnaires to the supervisors for clarification and they in turn would approach the data collector. These questionnaires were corrected and subsequently re-entered at the end of the field work.
Forms were devised by the supervisors to keep track of expenditure and to organise the field work. Three forms were devised;
b) One which recorded the data collectors transport costs.
c) A form for the supervisors that recorded where each questionnaire was, when they gave it out and when it was returned, and who filled it in.
The accounts were divided up between the three field offices with each office keeping their own accounts. They had been trained to use a computerised accounting system that greatly facilitated the control of the financial side of the survey.
They were then handed in to the in-field coordinator who joined all the accounts together.
The supervisors ran most aspects of the field work. They decided on the speed, which depended on the quality, they checked the questionnaires in the office before submitting them to the data entry and they had regular meeting to discuss the problems and to help clear up any misunderstandings as exceptions to the rules always occur. Approaches to difficult respondents were discussed and they were aided in their task by the use of a prepared leaflet that outlined the aims of the survey and gave a telephone number to call. Data collectors were also issued with ID cards, which they were required to show in each house.
Field work itself took a month (October 1993) to complete and this was followed by an evaluation firstly by the data collectors themselves, then a more thorough one by the supervisors jointly from Gaza and Jerusalem, which led to them writing a report summarising their recommendations.