The Art of Collecting Socio-Economic Data in Rural Areas* - An Example in Java
by Gunawan Wiradi**
Summary and Conclusions
The main topic in this paper concerns the problems encountered in the field while collecting data. These problems may seem unimportant and they are frequently neglected, but they can have a significant effect upon research results.
Among the problems of collecting data in rural Java are: the diversity of Javanese villages, the varying languages, and the problem of selecting field interviewers suitable for given locations. Other problems concern the time available and suitable for visiting the respondents, and closely related, is the way to behave towards rural people. In my experience, "low profile" attitude seemed proper. By this way of behaving, a good relationship with the people will be created, several biases will be reduced and the interview process, despite the need of much longer time, will yiedd reliable information.
1. Introduction
This simple paper is presented just to honor the honor given to me for giving a little contribution to this workshop.
I hope your understanding and forebearance if some of its contents do not meet your expectations. The shortcomings, and the simplicity of these explanations are merely due to the poor knowledge on my part of research methodology.
Theory and textbook are indispensable as a guiding tool for many kinds of problems. Yet since theory is mostly based on experiences, it is still true then, t hat experience is the best teacher. New experiences may provide, equip, and to some extent, improve existing theory.
It is only for that reason that I brace myself to write this paper, the contents of which are mostly based on my own personal experiences.
Recently computer analyses of large amounts of data has become important, and in general, social scientists are beginning to rely heavily upon them. Yet, sometimes some arguments arise because the results seem inadequate. Of course it is not the computer which is to be blamed. Most likely the cause of the irrelevant conclusion will be found in the reliability, the validity and the sufficiency of the data. In other words, the problem lies in how the data have been collected.
The study of rural societies is a voyage into a fascinating .strange land. Modern man cannot fully appreciate the human aspects of what he observes using the conventional instruments of the impartial observer alone. Unless the researcher is able to dive deep into the existential realities of the people, they are prone to make existential biases, irrelevant conclusions, and unwarranted premises. He would be much like Alice in her Wonderland.l/
Collecting data in rural areas will unavoidably communicate with the men who inhabit it. We are not only talking about rice, but the men who produce it. We are not merely calculating numbers and studying matters, but also observing the men who, unfortunately, differently interpret it. Unlike modern men for example who consider devil and deities as merely pseudo-images offering quasi explanation to the causal nexus of events, rural people may look at them as being no less real than airplanes.
So as soon as we are coming to the villages, we are directly confronted with the complexes of meanings. This is the world of norms and cherished values, of taboos, of emotions and affects, of beliefs and myth. This hidden dimension may not effect the validity of the derived conclusions and inferences if the researchers are studying a society similar to their own. But when they are trying to understand the meaningful relations in a society wholly different from their own, then some difficulties arise.
Field collection of socio-economic data in rural Java still needs those considerations because Javanese villages vary widely, from place to place. There might even be fundamental differences between two neighboring villages.
It is in this connection that I agree with the statement of Professor Srinivas that "if the social scientist really wants to know his rural area he must spend some period living in approximately the same conditions that the people of that area do",2/ because to understand social interrelationships and processes we need a basis for insights. In simple wording, participantobserver approach of the anthropologist is suitable. In practical operation however, to other social science disciplines, it might need some adjustments. But then, that is a problem.
* Presented in the workshop on "Field Collection of Socio-Economic Data in Developing Countries," held in Singapore, October 28 - November 2, 1974.
** The author is a s taff member of the Agro-Economic Survey of Indonesia.
1/ Hidajat Nataatmadja: "Rural Dynamics Methodological Framework: An Existential Point of View", Jakarta 1974. Economic and Finance in Indonesia, Vo1. XXII, No. 1, p. 78.
2/ M.N. Srinivas: "Village Living: A Source of Insights for the Social Scientist". Research Methodology (No. 35) A/D/C Teaching Forum, Singapore, January 1974.
2. Javanese villages and personal experiences
berore further discussing the problems in the field, let me explain first the conditions which any researcher in rural Java will have to put up with.
With regard to administration, Javanese villages are more or less autonomous. The headman, called Lurah, is an elected leader. In some places the secretary, called Carik or Jurutulis, is also an elected leader who functions as a second man, but in other villages, like the other members of the village government in general, he is appointed by the Lurah. This group of people which ranges in number from 10 to 20 (depending on the size of the village or on local custom), forms a governing body of the village people. Each member has a different assignment and responsibility (as village security, water distribution, message carrier, caller to prayer, etc.). They receive no salaries for these jobs. But as soon as they are appointed/elected, they have the right to operate the village land available and enritled for them. These lands, called bengkok, are distributed to them in accordance with their positions and responsibilities. So the Lurah has the largest bengkok.
Since the Lurah is responsible for his people and his village, he is usually the final decision maker on certain issues, though once in every 35 days there is a village meeting to discuss important subjects. This meeting, generally consisting of land right owners, is held principally to approve or disapprove of all decisions which have to be made. But in practice, especially in recent times, it is only asked for consultations.
Any stranger, foreign or native, who visits a village has to report to, and get the permission from, the Lurah, even after getting it from the higher authorities. The Lurah can even reject a person who is appointed by the central government (assigned to teach in the village elementary school, for example).
In most places, besides certain plots of land held by the Pamong Desa (The Village Government) as already mentioned, there is a plot of land called "bondo deso" which means village property. The Lurah controls this land, and usually its yield is specially reserved for extra expenses, such as accepting the village guests. Any researcher who stays for several days in the village will be treated as the village's guest and the Lurah will give him food and makes available accommodation for him without any payment. Sometimes, because of certain considerations, the researcher will be placed not in the headman's house, but somewhere else appointed by the Lurah. But this is an exception. However bad the conditions are, usually the researcher will be placed in the main house, in the verandah, or in a certain room apart from the main house, but still within the homeyard of the Lurah's house. Not only because by doing this the Lurah can make a direct observation on his guest, but also because he is supposed to be the only man competent (materially and responsibly)
for accepting respected visitors. Not doing that, he is afraid of being considered as disregarding the guest. Only once in my experience, was I placed in the house of one of the other village leaders, faraway from the Lurah's. I found out later that the reason was because the Lurah had a young wife.
It was in l960 that I had been staying in the village alone for a relatively long period, working on my graduate thesis, I had been sent to the village of Ngandagan in the district of Purworejo Central Java, for two months. It was really an invaluable experience for me.
I was accepted in a friendly manner by the Lurah and placed in a room in his main house. When I asked him whether or not I could stay in one of the villager's house, he refused. Feeling that it was unwise to argue with his refusal, I remained there for 60 days.
The loneliness of village life crept into my feelings on the first night I slept there. But surprisingly, the next day the Lurah called all the villagers for a public meeting. I was introduced to them although as soon as I stepped into the Lurah's house my name was spread within minutes throughout the village as being a man who will lead the people for developing the agricultural production and bringing up the village society. Trying to eliminate further misunderstanding, I stood up at the first opportunity given to me and said: "I am nothing but a student coming to this village just to check the correctness of the theories written in the books with the realities existing in the villages. In the classroom I may be your teacher, but in the field I am simply your foolish pupil!"
I didn't even imagine how this would please them. They gave me all of the information I needed. They explained about everything which they felt would be useful for me, without being asked. Well begun is half done. In the days that followed, my journey's path into the village study was smooth.
Every night I was accompanied by some village leaders, most often the Lurah himself, talking and discussing matters until late in the night. In addition, there were always two or three other villagers who came by turn each night to perform their duties as village watchman. I could not work well at night because of the lack of electric lights, but this was not the real trouble. It was the water that I really minded because my skin is very sensitive to alergic diseases. The bathroom and the lavatory were located outside the house.
During the first week I spent my time looking around the village, making friends, chattering with farmers, and walking along the rice field. But instead of walking alone, I was always accompanied by one of the Pamong Desa.
I didn't work with any questionnaires at that time, though I did have an outline as a guide for interviewing the villagers. Using an informal approach, I always kept the schedule in my pocket and asked them questions without writing any notes. Otherwise they would have been afraid and there would have been clumsy atmosphere. They were still not used to being asked formal questions using questionnaires. Anyone who did that, at least so they thought, must be a tax collector or special agent of the police.
Indeed, working in this way needed a good memory. Everytime I came back from interviewing, I directly wrote everything I could remember. In case there was something missing, I checked it the next day. I didn't limit the number of respondents but I looked for as many as possible until I was convinced that the information I obtained was sufficient. Besides checking with the other respondents, in order to reduce biases in the information made available by respondents, I always discussed it with Pak Lurah at night. And above all, I asked him to hold a small meeting with all the village leaders every week in which we could cross examine all the information.
The suitable time for visiting the respondents was twice a day, the first at some time around noon and the second in the evening. The farmers and the farm laborers usually worked in the rice field from 5 or 6 until 10 or 11 o'clock in the morning, after which they went home, took a rest and ate their meals. At 140 in the afternoon they went back to the field until 1700. After taking a bath they ate their evening meals at 1800, sitting and chatting with their family members for a while, and for the rest of their time, if there was nothing more to do, they slept. They had no breakfast except a cup of coffee, sometimes with some snacks. I had to get accustomed to the time schedule of eating meals, and I suffered from this in the first week.
Early in every morning when I was still sleepy, I was always awakened by the sound of loud arguments between women walking along the village road, especially on the marketing day. They were going to the market nearby which is, like any other Javanese village, a "five-daily" market. There they sold their produce or other items, and bought their necessities. If not to the nearby market, they went to the other ones which were located much farther than the first, on days according to the five-day cycle.
As I finished the field work and started to go home I tried to give some money to the Lurah, but surprisingly he refused angrily, and said: "If so, that means you didn't feel conveniently staying here. I assure you, we, the villagers did our best to make your stay amongest us comfortable. I'm very sorry, I don't like to accept the money!" I apologised for this. And the next day when I was really starting to go home, I was surprised by his giving me a train ticket, two bags full of eggs and other victuals, and a large basket of vegetables. I couldn't help refusing. So I went home, full of affection.
The villagers were really friendly if we knew the right way to get along with them. Despite the smll of cow-dungs and manure heaps and many other unpleasant realities of rural life, village living also means a feeling of tranquility and peace.
3. Choice of instrument and the problems in the field
In choosing the instrument for collecting data in rural Java, one thing is certain -- using mail questionnaire is impossible! It is appropriate only when the subjects have had a considerable amount of education. Most of the adult rural population of today attended elementary school only for three or four years.
In my opinion, a personal interview using an unstructured question format is preferable. Since my experience in 1960 mentioned above, the villagers are now getting accustomed to being interviewed with questionnaires, provided all they have to do is just to talk, not write. So now I usually structure the questionnaire, but not in too much detail, and frame the wording in simple sentences. But this is really not a standardized questionnaire.
Textbooks tell me that in the standardized interviews, questions should be presented with exactly the same wording and in the same order as written on the questionnaires to all respondents because otherwise the answers may not be comparable. But in rural Java it is impossible to practice that. The problems are first, the level of education (as already mentioned), and second, the language. The villagers in general are still unable to speak the Indonesian language perfectly and there are even a large number of people who can not speak it at all. Even the Javanese language is differentiated into many dialects and local styles so that a Javanese speaker might be unable to understand a statement made by those who come from a different area in Java. It is difficult to frame absolutely standardized questions which cover the entire survey area. The best way is, though the questions and their order are predetermined, for the interviewer to be given freedom not only to repeat the questions if the reply is not to the point, but also to ask any supplementary questions at his discretion. First of all, he must translate the questions into the local language. But even for those who can speak Indonesian, and even for closed questions, the interviewer must dig up and encourage the respondent to talk freely so that the reply will be reliable, otherwise there will be some funny answers. For example, if we ask: "yes or no?", the answer will almost certainly be "no"; and if we ask: "no or yes?" the answer will be "yes". Just like a learning child. Though the interviewer should not be too talkative, it is always necessary to explain the meaning of the question.
The other problem closely related to the problem of language is in selection of field interviewers. However short the time available, it is always necessary to train them first, or at least to accompany them on the spot as they begin interviewing the first four or five respondents. They must be prepared to recognize different local technical terms used in the village, especially for measuring the unit of weight, land size and the word for landrights. In most villages for example, one "dacin" means 62.5 kg, but in some other village it means l00 kg. An indifferent interviewer will usually overlook these problems.
The next problem is the limited time available for visiting the respondents. As I already mentioned, the suitable time is only twice a day, at some time around noon, and/or in the evening when the respondents are at home. It does not really matter for villages where the respondents are staying close to one another. But in large villages, where they live spread about at a distance, the fieldworker will lose much of his time if he fails to meet the respondents due to visiting them at the wrong time. Considering the available time and the contents of the questionnaires, I feel that 2 - 3 respondents per manday is adequate as a standard working volume. Depending on the Lurah, it is also possible to order all the respondents not to go to the field and stay at home at the time preferred by the interviewers. But even ordered by the Lurah, not all the respondents would obey. Some of them feel that working in the field or looking for money is more important than answering questions. The Lurah can also send for them to come to his house and the interviewer can question them one by one. But here, the situation will be too formal and consequently the respondents will have some feeling of fear. Hence, their responses might not be accurate. The best place to do the interview is therefore at their homes, in a friendly and free atmosphere.
In conducting a simultaneous quick survey, in which 50 to 100 sample villagers in every sample village have to be interviewed, it requires four to five field interviewers staying in each village for about one week. If they come from different institutions and have different backgrounds, their results may not be comparable because of their difference in skill and experience. To overcome this difficulty, we usually used a "group interview", which is held before the fieldworkers start to interview any respondents or after they had finished with all the respondents, or both. This forum consisted of all the interviewers, village leaders, progressive farmers, and if any, some educated villagers.
The advantage of it when held before any interviewing is done is that all the team will have a preliminary picture of the village problems. Some consensus amongst them will be reached, especially concerning the use of local terminology and unit of measurements. But the disadvantage is that some members of the team may be too much influenced by their preliminary pictures. On the contrary, when a group interview is held only after all interviews have been finished and evidently after being checked in this forum, the interviewers may have some difficulty in grasping the statements of the respondents, and may sometimes have to come back to their respondents. Therefore holding both group interviews will be more proper.
Besides that, it is necessary to give additional assignments to the enumerators. They have to make some report beyond the completed schedules or interview reports. Two advantages will be gained from it. First, being given this particular asaignment, the fieldworkers will be serioualy doing their jobs, not simply enumerating mechanically as it causes them to dig for reliable and sufficient information on certain problems. Otherwise they can write nothing. The second advantage is, there may be some special problem which might be overlooked by one, but not by the other members.
Field diaries are very useful, not only for the interviewers doing the additional reporting, but also for the researcher.
In supplying the field interviewers' equipment, we usually have to give them pencils, erasers, sharpeners and diaries. Each diary contains certain pre-determined items which will be discussed in the group interviews, and some empty blanks. These diaries must be returned later to the researcher so he can cross examine the interviewers and may find some special information related to the responses written in the schedules, or something entirely new.
4. Adjustments of behaviour and attitude in the field
The art of collecting field data in developing countries, or at least in rural Java, is unique. It needs some adjustments of our own behaviour and attitude, otherwise it will be difficult to establish good rapport with the villagers. In regard to foreigners, fortunately the villagers do not really mind about it. But when it involved the native interviewers, then the matter really needs attention. The villagers think that it is understandable if foreigners do not know about village customs, but for the natives, this problem may sometimes have serious consequences.
My general principle is, we go to the village to look for, not to give information. So, do not lecture the villagers at all, unless, on certain favorable occasions, they really ask for it.
In my experience it happened that several times this mistakehas been made. One person, immediately after arriving at the village, while visiting the first respondent, lectured him about farming technique. He felt that it would be useful. Unfortunately, the respondent was a relatively educated man who was apparently active in other occupations besides being only a farmer. After this lecture, he didn't respond to the questions at all and he did not want to meet my friend anymore. According to the other villagers, he said: "It is natural that he (the interviewer) knows more than me about farming, because he is a graduate, but I think I know more than him in the field, If he would like to ask me questions, just ask! Doing work in the field or looking for money is more important than merely hearing lectures! I have lost much of my time."
I usually took a certain attitude which I may call a "low profile" attitude, in order to establish a friendly atmosphere and good relationships with the villagers. We come to the village not as officials nor as extension agents or other authorities giving orders or lectures, but as the "needer" trying to learn something from the "needed." By taking a low profile attitude, some biases of courtesy and fear will be reduced or even eliminated.
Giving incentives to the respondents is not always necessary, though I always did, depending on the situation and conditions of the respondents. One pack of cigarettes is sufficient except for poor villagers to whom I gave some money. But even the poorest one sometimes refused to be given money especially when it was given in the wrong way. So, more important is the way the incentive is given.
Incentives to the village leaders are always necessary, at least in the form of cigarettes or victuals. Certainly, money is preferable but again here, the way to give it is important. If it is put into an envelope so that no one could see it, it will be felt more polite. We have to give them not because we are generous nor they are really expecting it, but simply because otherwise they will consider us as being a man who does not know how to thank them for their service. Unlike my experience in 1960 when the Lurah refused to being given some money, money incentives are now welcomed. Despite having certain land, the yield of which is specially used for accepting guests, he really needs money. Giving food to a team of several members staying for several days is really a heavy burden for him. It is only his sense of dignity that makes him not ask.
Back to my principle, taking a low profile attitude which will eliminate certain images such as that of a judge questioning an accused, or a tax collector investigating the people's property.
When we interview the villagers at their homes, first we have to creaa friendly atmosphere by giving some introduction. But unlike in developed countries where just saying "Good morning" and showing an identity card is enough, in rural Java it requires a much longer time. Indicating identity cards is not necessary, because it may even create a sense of fear. Tolerance and forebearance will be needed, also during asking questions. It happened sometimes that my friend who accompanied me interviewing an illiterate old respondent felt asleep because it was too exhausting digging up the reliable informations.
After my long experience in collecting field data in rural areas, I am almost sure now that if the interviewer is in too much of a hurry in doing his interview, the data collected will be unreliable, even if it is quantitative objective data. A hasty manner and a clumsy attitude will make the respondent give haphazard answers.
5. Reducing or adjusting some biases
Besides biases of courtesy and fear, the other ones due to limitation on recall or non response were common. Not being used to recording their own activities, the villagers usually forget what had been done, or happened in the past, even for events which happened only a year before. To refresh their memories I usually took some popular important event as a directive "milestone", such as the fast month, independence day, or something else important. Another possible way is to check backward from now, month by month or even week by week. That is why interviewing the villagers really needs much time.
To verify the responses in the field, several ways could be taken. For personal activities, in my experience, the most useful is to check within schedules. So we have to frame the questionnaire in such a way that cross checking within the schedule is possible. Different questions, though closely related to each other must be put on different pages sometimes not in succession Successive ordering is ideal, but for some problems which need cross checking it might lead to biased answers. The answer of the subsequent question might be much influenced by the former one. For instance, we asked about the total amount of rice sold during a given season, then we asked about the selling price month by month. Successive ordering might lead the respondent to frame his answer in such a way that all figures will come into place. If only the first answer were reliable.
For objective data like the size of land or the number of houses owned, etc. it is always better to check with the village leaders, or if any, looking at their certificates.
With regard to non-response there is another way to reduce biases. Facing a silent respondent I usually closed my questionnaire, relaxed, and talked to him about everything informally. About family, music, about his hobbies or anything else until he was interested. Then I let him talk
freely, and sometimes interrupted him with non directive questions until the information I needed was forthcoming.
Non response is usually due to fear or suspicion. Here again, the manner in which the interview is performed is important.
Above all, besides external objective data, checking with the village leaders in a group interview is very advantageous.