8.5 Embebimientos
9.2.3 Productos y espacios de Hausdorff
The research population for this study consisted of all native speaking Russian individuals over 18 years, and who predominantly (see below) legally reside on the territory of Russian Federation. The unit of analysis was the individual respondent.
Native speaking Russian residents of Russia referred to members of ethnic groups officially residing in the Russian Federation after the dissolution of USSR in 199119. Temporary visitors of Russia, even if they spoke the Russian language as one of the native languages, were not included in the research population. Also illegal aliens, who reside in Russia without official registration, were not included in the research population. “Predominantly” was used as legally defined in the Tax Code of Russian
19 -this includes Russians and 47 other ethnic groups, territories of whom were officially included in
the “Russian Federation” as of 1991. Ethnic groups that were part of Russia prior to 1991 and terminated their memberships in Russian Federation on 1991 were no longer considered members of the research population
Federation, article 265: a person who cumulatively resides on the territory of Russian Federation not less then 260 days per one calendar year. The research subjects had to reach their 18th birthday at the time of the interview to be considered a member of the research population. The screening questions were used in the beginning of each interview to select the appropriate research subjects from the pool of potential candidates.
4.2. Sample Size
The sample size used for this study is equal to 1600 individuals. The sample size was estimated using a 95% confidence level, 5% sampling error, a medium size effect, 0.8 statistical power, and a known study population size of 142,800.00 as reported in the 2006 Russian Official Census data (Sproull, 1995). The choice of statistical analysis method (structural equation modeling) and the total number of variables (forty-four) were also taken in to account. The literature suggested the use of at least ten cases per parameter to ensure a sufficient level of statistical power in the SEM research (Bentler, 2004). Holster’s Critical N (CN) is also employed to evaluate the adequacy of the sample size for the structural model used (Maruyama, 1998).
4.3. Sampling Methodology
For the purposes of this study a random stratified cluster multistage sampling method was used. This sampling technique was employed to produce a representative sample of the entire adult Russian population.
In the first stage of sampling, clusters of regions of the Russian Federation were being randomly selected. In 2007 Russia consisted of 89 regions including those that are called “republics, areas and federal cities.” According to Russian federal law the 89 regions are geographically grouped into the larger fractions called “federal circuits.” There are seven federal circuits in Russia. To do the sampling selection, regions within each of seven federal circuits were grouped into the clusters. These clusters were organized based on the following known characteristics of the regions:
1. climate conditions 2. the density of population
3. the proportion of urban to rural population 4. the average level of income per capita
5. the level of transportation infrastructure development 6. the type of industry developed in the region
All 89 regions were grouped into 44 relatively homogenous clusters within seven federal districts. From the list 22 clusters were randomly selected for the sampling. These clusters consisted of 50 regions of Russia.
In the second stage of sampling the cities, towns, and villages were selected from the clustered regions. The following strata were developed: super-large cities with population over one million people; large cities with population from 500 to 999 thousand people; middle-sized cities with populations between100 to 499 thousand people; towns and villages with a population of less than 100 thousand people. From each of 22 selected clusters at least one, and no more than four cities, towns, or villages were randomly selected for each of the five strata. The overall number of
cities, towns, or villages selected in this stage within each circuit, and within each stratum, was controlled for the known distribution of the population in the Russian Federation. The number of cities, towns, and villages selected from each circuit reflected the overall density of the population in each region. Highly populated circuits like the Central, Volga, and Southern20 had high numbers of cities, towns, and villages sampled. The de-populated circuits such as North-Western and Far East circuits21 received the smallest number of selected cities, towns, and villages. The strata “villages” and “small cities” received the largest number of selected items, because according to the Russian Census data, 26.6% and 26.2% of Russian population respectively was living in the villages and small cities. Only 11% and 9.8% of Russian population were living in the super-large and large cities respectively, therefore these strata received the smallest number of selected items. The number of villages was over-represented to exclude the possibility of under- sampling in a small geographic location (within one-two blocks).
Table 19. Number of cities, towns, and villages selected in each circuit
Circuits/Strata Over 1 million of people 500-999 thousand of people 100-499 thousand of people Less than 100 thousand of people Villages Total Central 1 2 6 5 5 19 North-West 2 0 3 3 4 9 Volga 5 5 4 3 6 22 Southern 2 1 5 4 6 18 Ural 2 1 2 3 2 10 Siberian 2 2 4 3 4 15 Far East 0 2 2 2 2 8 Total 14 13 26 23 29 101
20 - over 90% of regions within these circuits have population over 3 million people according to
Russian censor data for 2006
21 - over 90% of regions within these circuits have population ranging from 500 thousand people to 1
In the third stage of the sampling, the election districts were selected from the list of cities, towns, and villages. The election districts are official territorial divisions of Russia within particular city, towns, and villages that are being used to organize all local and federal elections. Each election district has an identification number and a known number of registered voters. The number of registered voters per election district usually ranges from 100 to 1000 people. The election districts were randomly selected with the probability proportionate to the number of registered voters. In each selected city, town, and village three election districts were sampled. In villages with less than three election districts, all available districts were included in the sampling. Thus, on the third stage, the sample consisted of 289 election circuits.
In the fourth stage of sampling, the household registered with each election district were selected. This selection was done by a simple random method where each household had an equal probability of being selected in the sample. Due to a possible non-response, the actual number of households selected was 25% higher then the sample size. The total number of selected household on this stage was 2000.
In the fifth stage of the sampling, the individual respondents were being selected within the household. The selection was done using the quota requirements contained in the task list for each participating interviewer. The quotas were constructed based on the age and gender parameters. The known age and gender distribution within each region was employed to create the quota task for each interviewer. The interviewer was not allowed to interview more than one person from each household, and was allowed to use no more than three households from one building.
4.4. Sampling Frame
Different sampling frames were used in each of the five stages of sampling selection. For the first stage of sampling the list of 89 regions in Russia listed in the Russian Constitution of 1993 was used. For the second stage of sampling, the official list of cities, towns, and villages within the region registered by the regional administration for 2006 was used. The data regarding the size of the population within each city, town, and village was taken from the 2006 Russian Census Data collected by Federal Agency of State Statistics of Russia (RAS, 2006). For the third sampling stage the official list of election districts registered within each city, town, and village by the Central Election Commission of Russia was used. In the fourth stage of sampling multiple sampling frames were used for each election district. In each district the official list of registered households was used. In the fifth stage of sampling the list of household members, who were currently considered registered voters with the election district, was used. This list was constructed by each interviewer after the particular households had been selected. The registered voters in Russia are citizens that are 18 years old at the time of registration with the election district.
5. Survey Administration and Data Collection