Survey of Conflict Prevention and Cooperation 2004
The project uses public opinion polling to gather and then analyze a sample that represents the entire population of each of four different countries of Central Asia: Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan.
Kind of Data
Sample survey data [ssd]
Data provided to the World Bank by The Brookings Institution on January 31, 2006
The project uses public opinion polling to gather and then analyze a sample that represents the entire population of the country.
Producers and sponsors
The Brookings Institution
For all four Central Asian countries in this survey, the sampling procedure is a three-stage stratified clustered one. Census data on the territorial dispersion of the population is used as the base to start the sampling methodology. The sampling procedure takes the total population of the country, considers geographic units within the country as either urban or rural, and then develops random procedures to select who to survey in three stages: first by randomly selected smaller geographic urban and units in each province (the primary sampling units or PSUs), second randomly chosing households within these units, and third, to randomly select which household member to interview in each household.
The sampling frame used to divide these four countries into smaller geographic units to randomly sample from differs slightly for each Central Asian country, based on differences in data availability on the population of the country and its dispersion. Subsequent sections explain the sampling methodology used and how this sampling frame differs in each country. Then all four countries have PSUs, random selection of households, and random sampling of individuals within households using the same methods.
Uzbekistan has 12 provinces, the Republic of Karakalpakstan, and the city of Tashkent. Each province has several districts for a total of 168 districts in the country. Each district has a number of cities, small towns and villages. Of the 233 cities and small towns in Uzbekistan, 76 cities are subordinated directly to provinces due to their importance. The population of Uzbekistan was 25,523,000 people, of which 9,410,700 (37%) were urban residents, and the 16,112,300 (63%) were rural residents as of May 2002. Several districts, practically inaccessible from an absence of transportation or remote location, are excluded from the sampling frame. These two cities, one small town, and one district in Navoi have a population of 95,300, 0.9% of the urban population and 0.1% of the rural population of the country - a total of 0.4% of the population of Uzbekistan is excluded from the sampling frame.
The sampling frame for Uzbekistan has primary sampling units (PSUs) of two types:
- MK ("Mahallinskiy Komitet") - town makhalla committee. Makhallas are the traditional neighborhood committees which have been revived (and in some urban areas artificially created) by the Uzbek government;
- SSG ("Selskiy Skhod Grazhdan") - village council. This type has been used for rural areas in all recent surveys.
The sampling scheme then has the following three standard stages:
- proportionate stratification by population of provinces;
- for all provinces (include Tashkent city as urban stratum):
- proportionate stratification by urban/rural population within provinces;
- PPS-sampling of PSUs within urban/rural strata;
- sequential random sampling of households (Secondary Sampling Units - SSUs) in selected PSUs;
- Kish grid based sampling of respondents.
Thus, the sampling is three-stage stratified clustered sampling.
There are 63 PSUs are selected from the sampling frames, with the number of respondents to be interviewed in each varying between 17 and 29 in different PSUs.
The sample distribution by the main demographic characteristics can be compared with data of Statistical Department of Republic of Uzbekistan from January 1, 2002.
During the fieldwork, 766 cases of non-response were registered (non-eligible units are excluded from this count). The average response rate is about 66% (1,500 of 2,266 attempts). Generally, the non-response case was registered if an interviewer had made up to two failed callbacks.
the response rate in rural areas is higher than in urban areas. In Tashkent city very much high level of refusals is observed (response rate barely about 38%). This is caused mainly by the following factors:
a) rural residents are more willing to cooperate;
b) they are less active in sense of movement, therefore more reachable;
c) the theme of interview sets people on the alert;
d) population registration and register maintenance in cities are generally worse which leads to poor quality sampling frames.
The influence of first two factors is aligned lately because of a falling of a scale of living of people.
40% of all the causes in the urban areas is the "household members refused contacting respondent" (cause 7), as compared with the corresponding 31.2% in the rural areas. This cause has the most spread for urban people and the second at the prevalence for rural areas (about 31% of all causes of non-response), because the theme of interview (the internal politic, interethnic problem etc.) makes people mistrustful and situation with the criminality (especially in the cities) is very complicated.
Otherwise, cause 10 ("not at home for a long time") is second at the prevalence for urban areas (about 37%) and first for rural areas (about 39% of all non-response causes). This cause is spread for urban and rural people because they migrate in searches of earnings.
The similar reasons called cause 3 "nobody at home" and 4 "respondent was not at home by that time" (8.2% and 2,3% for urban and 5.1% and 3.6% for rural areas accordingly). Besides for these causes there is one more explanation - employment of urban population and "cotton campaign" for rural population.
The causes 6, 8, and 9 met not frequently. Therefore we may not make any conclusions.
The sampling frame quality is revealed by comparing the share of cause 11 "address was not found, does not exist"- 4.8% in the urban areas versus 6.4% in the rural. In the urban areas 2.8% of the non-response are "Address is not residential" (cause 12). In the rural areas this cause makes 4.2% of all causes of non-response. In most cases it originates from that a household, in order to get an additional land plot from a makhalla committee for running subsidiary economy, declares itself to be actually consisting of two households - parents' and a new, young one. Then the makhalla committee registers a new household and allocates a plot. However, this "household" continues living with the parents, making the new address not residential. Most urban cases are connected with fitting apartments for small offices, cafes, renting to foreigners, etc. More apartments in the cities are thrown (owners have left in searching of earnings).
Dates of Data Collection
Data Collection Mode
Field works were completed (latest questionnaires entering) on 21 September 2004. Processing (check, codification) were started immediately upon the first questionnaires entering, i.e. since 11 September 2004. The first questionnaires were entered into the database on 14 September 2004.
Quality control included the following stages:
1. Check of correctness of the sampling filled in forms.
2. Check of correctness of respondent's sampling.
3. Check of the correspondence between the number of Kish' card and the number of the sampling form.
4. Check of correctness of the questionnaires filled in.
5. Check of the correspondence between answers and questions in terms of logic.
Check visits were made from 23 September until 30 September 2004.
After the field works' completion each interviewer, who participated in the questioning, was checked. With this purpose 20% of the conducted interviews were selected, which were presented to the headquarters' as filled out questionnaires. An interview' selection for check was made at a random in the database from interviews, submitted by each interviewer. Addresses, where an interviewer had made mistakes while filling out during questioning, were added to them (there were about 30 such addresses). Interviewers, who worked in other regions, performed check.
If the check's results proved to be positive (the procedure of respondent's or household's sampling was not violated or questionnaire was filled in according to the interview taken etc.), the check of the given interviewer was considered completed. If an interviewer under check made a sampling mistake or falsified the interview, than all the interviews, he had conducted, were subjected to check.
The following violations were discovered:
- In Samarkandksaya oblast in PSU No 32 (the town of Kattakurgan), due to the illness of an interviewer, a supervisor had to urgently replace him with another one, and so an interviewer, who had a bad command of Uzbek language, was performing interviews in Russian. In this connection we decided to send another interviewer so that she interviewed those respondents, who either did not speak Russian, or spoke it no good, or understood poorly at least one of the questionnaire's item. Such respondents were three in number and another interviewer questioned them.
- In the Republic of Karakalpakstan one case of the respondent's replacement was found out and the checking person conducted a repeated questioning.
- In Navoiyskaya oblast two cases of the respondent's replacement were found out, but it was discovered during the questionnaire's hand over and the interviewer himself performed a repeated questioning. No remarks were made during check regarding his other addressees.
- In Bukharskaya oblast 1 case of the respondent's replacement was found out, and the checking person conducted a repeated questioning.
- In Andijanskaya and Ferghanskaya oblast 1 case of the respondent's replacement per each were discovered as well, and an interviewer in charge of check conducted a repeated questioning.
- In Surhandar'yinskaya oblast - 2 cases.
- And in Tashkentskaya oblast -2 cases.
Totally 10 respondents' replacement were exposed, where they were re-questioned by other interviewers. It should be noticed that these cases of respondents' replacement were not deliberate, but owing to a technical error of the interviewer either at the respondent's sampling (3 cases) or at compiling a household members list (breach of men and women's age priority).
In the course of check visits one case of an interview's falsification was found out (PSU No 28), i.e. "non-visit". The interviewee was a 97 years old woman, but during check she said that "the interviewer came, but asked only just about 20 questions and left". During check a repeated questioning was made at this address too.
We have carried out check visits to these households as well, where for whatever grounds an interview had not been conducted. In the course of checks many cases were revealed when inhabitants either do not agree to be interviewed or they do not live in their apartments. Many apartments according to explanations are empty as they are for instance let to students, who have gone to gather cotton, or they are purchased as real estate etc. Here are concrete examples of such checks:
The town of Andijan
- In PSU No6, neighbours - household 131 - said that a young family moved in 2-3 days ago, but they did not live there yet.
- In household 130 - "That period I had lived with my son in another place (another district). The neighbours told me that somebody came and asked for us".
- In household 129 - "Yes, they came, but we did not want to converse, as I have a sick 75-year old mother".
The town of Namangan
- In household 593 - "Students live there, but presently they are at cotton gathering works". Neighbours.
- In household 591 - "We have bought it recently, and now making repair". " The old owner also used to come here seldom, he was a businessman". Neighbours.
- In household 587 - "Inhabitants show up sometimes, but do not live here". Neighbours from the upper floor.
The town of Kuvasai
- Household 1133 - "Dwellers live here, but come very late and leave very early, as they work at the plant. That woman (interviewer) came three times, but failed to find them home". Neighbour.
- Household 1134 - "They do not live here, for they have a village house. Mahalla representative.
- Household 1138 - "They do not live here and they have even been sued for municipal services non-payment". Neighbours.
- Household 1140 - "Few people live here, as gas, water etc. are unavailable". Neighbours from apartments number 17, 18.
- Household 1139 - The house is situated in the corner of the crossing and according to mahalla lists, it is referred to Kodiri street (house 14), but virtually its dwellers say, that their postal address is Burhonova street 19. They also said that a part of house is referred to Kodiri street till now, but it's a non-living area.
- Household 1139 (Azizova street, house 203, apartement 41) - "The owner of apartment shows up sometimes, but he lives in another district, in a land plot house". Neighbour Faya from apartment 42.
In the city of Tashkent under various reasons people in charge of check could not find the dwellers neither and got an answer from neighbours:
- Household 1464 - they would not open the door and shout at the controlling person to get away. Evidently, the interviewer had been treated in the same manner.
- Household 1463 - here lives Feruza, but she is working days and nights (in "Ambulance". Her son Timur and someone else were at home, but they did not open the door.
- Household 1449 - A dweller of this house Mukaddas is in the hospital.
- Household 1456 - This apartment has been in repair already for 3 weeks, and maintenance workers answered that the owners live in another place.
PROBLEMS DURING QUESTIONING CONDUCT
Different problems arose during questioning conduct. For instance, in Samarkandskaya oblast local authorities did not permit to conduct survey until superior (oblast hokimiyat) authorities endorse it in written. It usually takes 3-4 days to obtain such permission (the time, this issue is studied in Khokimiyat). But in this case the supervisor phoned directly from the Khokimiyat's representative and afterwards a promise was given to accelerate the process. Works were started the following day.
At PSU 45 of Angren town there were problems, i.e. the mahalla committee's representative refused to grant households lists without the authorities (Khokimiyat) permission. But after the interviewer went to khokimiyat with a letter and showed it to them, the khokimiyat representative made a call to the mahalla committee telling them to render assistance to her.
In Tashkent City almost in all (in four among five) mahallas there were similar problems. There were difficulties in interviewers' contacts with respondents. For instance in Tashkent, when an interviewer knocked the door, this household's dwellers told him to wait and asked by telephone their neighbours to come. When the neighbours gathered around, the dweller said that he did not know this individual (interviewer), but that the latter would like to visit them.
In many places, mostly, in rural locations, interviewers had to work from early morning until late evening due to cotton gathering season. People went to fields early in the morning (since 5-6 AM) and came back home towards 21-22 PM. That is why interviewers had also to visit them at such hours.
Interviewers had some difficulties in obtaining household lists of a PSU, where PSU consisted of a few settlements. For example, in PSU No33 an interviewer had to collect lists of all three villages, as not all lists were in SGB.
In PSU No2 due to the household lists unavailability in the mahalla committee (as they had been handed over to house committees for update) an interviewer tried to collect these lists from these house committees, but without success, as they, in their turn, also handed over the lists to their assistants. So the interviewer had to find the mahalla' s map, where all mahalla's street were designated and houses numerated, and to make sampling.
In Kashkadar'ya oblast, Guzar district, our interviewer asked a woman from this village to show locations of several households, and she joined our interviewer for half a day. But the inhabitants of PSU No21 took our interviewer for a wahhabist and wanted to drive away from the village the woman, who assisted her in showing addresses, for "the help to a wahhabist". On the second day our interviewer was detained by police and after examining her documents, they warned her, that the mahalla's women were waiting to meet her in the street and even to beat her. As to the woman, who showed her addresses the first day, she was already "done", i.e. they came to her house, scolded and insulted her, etc. and left only after her husband came home. The interviewer had to ask a police station officer that somebody from the mahalla's committee accompany her. For this reason she had many refusals in this village, (the rumours of the wahhabist' coming to their mahalla ran ahead of the "wahhabist" (our interviewer) herself), and so the inhabitants feared to answer her and would not open the door or pretended that they were out).
According to observations and notes of interviewers during questioning, various obstacles impeded their work. So, rather frequently during interviews, the interviewee's relatives would also announce their wish to participate in them. Most often it was done by men (husbands or fathers), who did not want to leave their wives or daughters all alone with the interviewer. In the majority of cases, interviewers, using skills, obtained during contact establishing training and the interviewer's behaviour in likewise cases, managed to persuade other family members not to distract the interviewee, but sometimes it was rather hard. Here are some observations of interviewers, which complete the questioning situation:
The respondent's (woman's) husband did not permit to interview his wife, explaining that they were not in need of additional problems. But in the given case, the interviewer asked to call for both his wife and other relatives. When they came up, he once again made the purpose of his visit and the survey clear to them and said, that if they did not agree to be interviewed, he would have to stay for a few days more to find somebody to replace them. Afterwards they gave permission. But during the interview the respondent's husband did not leave them alone and did not go away. According to the interviewer, though, he absolutely did not interfere and did not join the interview. Or others would say "And what shall we get? What for should we waste our time?" In such cases people did not mean personal profit, but the benefit for all, and they asked if this survey would promote to "plants and factories opening in their settlement or district" or "if the researchers (we) are able of augmenting their pensions" etc.
Interviewers had to "talk into" not only ordinary people, but even mahalla representatives. For example, a similar situation occurred in a mahalla, when its representative said to the interviewer that "it is roughly tens of times that likewise organisations with various surveys come to us, but not a single one has brought changes for the better, which we might feel". It took place in PSU ?36 of Samarkandskaya oblast. The interviewer responded to him, that "even if our results be used for improvement in any sphere, it is will not happen that soon, as before and after materials are submitted, it takes a few years for consideration of issues. And we are just researchers, our goal is to study public opinion regarding this or that matter, and other people, experts, are in charge of improvement, so we are unaware when our survey will be used for improvement or the issue solution".
Many respondents were interested to know what the use, they would get from those questioning, is, and if their lives would improve afterwards. Some relatives put pressure on respondents, trying to persuade them not to waste the time. Most often the respondents' behaviour was normal.
In general, 2 kinds of people's opinions are most often spread - that any conducted questioning is aimed at the life's improvement and that it will give its results. The second one is the incredulity that any benefit may arise from these surveys. This is the source of their inadequate reaction towards interviewers, accusing them in doing waste job, which brings benefit to nobody. According to the interviewer, there were instances, when even mahalla's representative told him that it was useless and asked why he (interviewer) does not fill in questionnaires and hand over his work instead of visiting households. And that he ostensibly knows that the population would fail to answer properly in any case. Here the interviewer explained that "the given questioning is a sampling study of the population opinion on living standards, national economy situation and the ties with other Central Asian countries and that each respondent's opinion is very precious to us and we need opinions of many individuals. And that we have no aim to get "proper" answers, but any people's view, and therefore we do not divide the responses and opinions into right or wrong".
Though the interviewers had passed "the interview conduct technique and the correct contact establishing with the household members" training and they had on their hands "general guidance on an interview conduct", many of them, who worked in town, received refusals. The explanation lies in rural inhabitants being more amiable, hospitable, as compared with their urban counterparts, and one can always find another member of the household to talk to, identify a respondent and appoint another hour of visit or ask him to inform the respondent about it. As for cities, it is impossible there, as children are at school, parents are at work, and the elderly will not open the door. Besides, in cities people are busy from morning till evening and after work they may refuse to grant an interview, whereas in villages even after work people usually do not to refuse the guest.
Data Collection Notes
The piloting of the questionnaire was performed from 13 august until 15 august. To conduct the pilot survey of 60 questionnaires, 3 mahallas were selected - one in Tashkent City, the second one in the village of Ulugbek and the third one is the village of Chinabad of Kibrai region. Skilful interviewers and supervisors, which have experience in questionnaire conduct under various themes and who have worked as a minimum 2-3 years in relevant field, were selected to perform questionnaire piloting.
Training for piloting questionnaire was performed on 13 August 2004. Goals and tasks of the questionnaire were explained, analysis of each question from the questionnaire was made, a working schedule and the schedule of the survey conduct were discussed at the training. Representatives of the Client were present at the training. The Centre Director and Project-Manager carried out the training. The Clients' representatives also took part in interviews at places with our interviewers on 14 and 15 August.
On 17 August the outcome of the piloting survey was discussed with participation of all the interviewers and client's representatives. During discussion all the issues, which revealed respondents' difficulties or problems in understanding, were examined. Under the piloting outcome many amendments were added to Russian, as well as to Uzbek and Karakalpak variants of questionnaires.
TRAINING FOR SUPERVISORS AND INTERVIEWERS
The following types of training were performed during preparation for the questioning:
- Training for piloting survey was carried out on 14 August 2004.
- Principal training for Tashkent interviewers was held on 4 and 5 September 2004.
- Training on the sampling and identification of the respondent under Kish's card for new interviewers, who took part in similar surveys on 6 September 2004.
- Training in oblast centres for local interviewers - depending on the beginning of field works in each province (from 7 September up to 9 September 2004).
(1) Training for supervisors and interviewers.
So that interviewers perceive goals and tasks of the survey at a more qualitative level, the training was divided for 2 days held in Tashkent and conducted by the Client's Representative and Project Supervisor, and consisted of the following procedures:
- Training of interviewers and supervisors on interview conduct guidance. During this training a tête-à-tête interview with interviewers and exemplary performances, imitating the establishing of contact with households and respondents were played.
- Explanation of the general survey strategy.
- Explanation of respondents' sampling rules for the survey.
- Explanation of households' sampling rules for the survey according to the sample form.
- General interviewing face-to-face questions and the technique of an interview performance.
- A question by question analysis of the questionnaire.
- Conduct of trial interviews.
- Analysis of field works organisational issues in oblasts.
- Sampling of interviewers.
Practical lessons on performance of sampling at places of questioning, as well as trial interviews between interviewers were carried out along with the explanation of general survey strategy, rights and responsibilities of supervisors, analysis of the questionnaire and discussion of some organisational aspects.
(3) Training for interviewers in provinces (oblasts)
Training for interviewers in provinces was carried out by provincial supervisors depending on the beginning of field works in every province. All in all 56 interviewers and 7 supervisors participated in the survey. Among them 48 interviewers speak Uzbek and Russian languages, whereas 3 interviewers speak only Russian and 4 interviewers - only Russian and Karakalpak and 1 interviewer - Russian, Karakalpak and Uzbek.
REGIONAL DISTRIBUTION OF INTERVIEWERS ACCORDING TO THE AMOUNT OF CONDUCTED INTERVIEWS
While distributing the interviews' amount between interviewers, we adhered to the principle that each interviewer should have no more than 30 interviews. Still in some cases this principle had to be broken. This was stipulated by different efficiencies of interviewers, as well as by some specifics in the questioning' s organisation and by the fact that prior and after questioning in provinces, some interviewers had to work in Tashkent. Besides, households' repeated questioning was to be made after control visits to some locations. Supervisors, who organised field works in provinces, were not prohibited to conduct interviews on their own, if there were no obstacles in actual situation.
REPEATED VISITS AND REFUSALS TO PARTICIPATE IN AN INTERVIEW
In accordance with specifications, if an interview performance was impossible at the first visit to the household, an interviewer was to make not less than 2 repeated visits. Only after the third visit, when it was still impossible to carry out an interview, did the interviewer stop his efforts and registered his visit as a failed interview, and switched to another addressee. As a rule, interviewers made repeated visits the next day. Sometimes, however, a repeated visit took place at another hour of the same day (for example, the first visit was in the morning, and the second one in the evening).
Reasons of refusals to grant interviews during field works were different.
According to interviewers, mostly (in most cases) it was caused by the absence of household members at the given address. If in town this is due to urban dwellers being at work the whole day, in rural location it is owing to cotton gathering season. It is for this reason that interviewers had to work since early morning (before people go in the field) until the latest hour.
The second ground was a household member's or a respondent's firm refusal from an interview. Obviously this refusal is based on the incapability of interviewers to persuade a required respondent to agree to converse, so in the future this aspect of questioning performance should be attended to more scrupulously. The same refers to the refusal of other members of households to contact a required respondent. These reasons are a specific urban phenomenon, that is why interviewers, who work in town, should receive further particularly meticulous instructions before starting questionnaire.
To perform questioning, the following documents have been prepared (attached):
- Questionnaire (in Uzbek, in Russian and in Karakalpak languages).
- Sets of cards (in Uzbek, in Russian and in Karakalpak languages).
- Forms of the respondent's sampling and records of the households' visits with Kish's cards (in Russian and Uzbek languages).
- Forms of the households' sampling in selected points of questioning (in Russian language).
- Sampling instructions (in Russian and Uzbek languages).
- Instructions on households and respondents' sampling (in Russian and Uzbek languages).
- Examples how to fill in sampling forms
- Covering letter to local authorities of 2 types (in Cyrillic and in Latin).
- The data and related materials will not be redistributed or sold to other individuals, institutions, or organizations without the written agreement of the Microdata Library.
- The data will be used for statistical and scientific research purposes only. They will be used solely for reporting of aggregated information, and not for investigation of specific individuals or organizations.
- No attempt will be made to re-identify respondents, and no use will be made of the identity of any person or establishment discovered inadvertently. Any such discovery would immediately be reported to the Microdata Library.
- No attempt will be made to produce links among datasets provided by the Microdata Library, or among data from the Microdata Library and other datasets that could identify individuals or organizations.
- Any books, articles, conference papers, theses, dissertations, reports, or other publications that employ data obtained from the Microdata Library will cite the source of data in accordance with the Citation Requirement provided with the dataset.
- An electronic copy of all reports and publications based on the requested data will be sent to the Microdata Library.
- The original collector of the data, the Microdata Library, and the relevant funding agencies bear no responsibility for use of the data or for interpretations or inferences based upon such uses.
Uzbekistan Survey of Conflict Prevention and Cooperation 2004, Ref. KAZ_2004_SCPC_v01_M, dataset downloaded from microdata.worldbank.org on [date]