Põlvkondade eluteed 91
Respondents Characteristics of the respondents in the Parental Cohort by Gender, Education, Age, and Nationality Estonians Non-Estonians Number Percent Number Percent Gender Male 477 34.0 199 38.3 Female 925 66.0 320 61.7 Education Higher (some or complete) 701 50.0 318 61.3 Secondary specialized 284 20.3 104 20.0 General secondary 417 29.7 97 18.6 Age 42 and younger 416 29.7 253 48.7 43 738 52.6 181 34.9 44 and elder 218 17.7 85 16.4 Table1 presents some general information about the respondents. There are specific reasons why respondents are different from the normal population distribution. Our main sample was based on the 1979 sample, which consisted almost entirely of ethnic Estonians (persons who were studying in Estonian language schools in 1966). Most of the non-Estonians were added to the survey only as control group. Partly for this reason, the analysis of data could not simply combine Estonians and non-Estonians together as if they were representative cross-section of the population at large. There is also some gender disproportion in our sample, because in the 1960s Estonia did not yet have compulsory secondary education and many young men (more than women) decided not to continue their education in secondary school after graduating from the eighth grade. The educational level of the respondents were also much higher than Estonia's population as a whole. Attaining secondary education was not very common in Estonia in the 1960s and more than half of secondary school graduates became students in higher educational institutions later that same year. Relatively small age difference between Estonians and non-Estonians were caused by fact that the Russian language schools were ten-year schools and their students graduated at age 17 while the Estonians graduated at age 18 from eleventh class.
The Estonian Longitudinal Survey (ELS) actually began in 1966 when Mikk Titma conducted the first stage of survey on a cohort of Estonian Young people, who were at that time eighteen-years-old students in their last year of Estonian-language general secondary school. Titma last interviewed these respondents in 1979. Working with an international team of researchers, Titma administered another round of the ELS between February and July 1991, while, at this stage, expanding the survey's scope. Following overview describes the specific methodology of the 1991 ELS, including the instrument and sample designs, the field procedures, and the characteristics of the respondents. A short history of the project The project's beginning was in 1966, when Mikk Titma conducted the first stage of survey on a cohort of Estonian young people. When interviewed for the first time, the cohort (born in 1948) consisted of students who were in their last year of general secondary schools in which Estonian was the main medium of instruction; almost all of them were ethnic Estonians. Titma subsequently reinterviewed this cohort several times: 1969, 1973, 1979 and the last time in 1991. As part of a binational (United States and USSR) commission on comparative longitudinal sociological research, Barbara A. Anderson, Brian D. Silver, and Mikk Titma began a collaborative study of the life course of youth in Estonia, the Soviet Union, and United States in 1988. This international collaboration, set the stage for the last round of the survey, which was carried out between February and July of 1991. This stage of the Estonian Longitudinal Survey (ELS) expanded its scope. In addition to reinterviewing the Estonian respondents from 1979, the new survey incorporated a control group from the same cohort who had completed general secondary school in Russian-language schools in Estonia 1965 and 1966. Most of these respondents were self-identified Russians, but many were Ukrainians, and some were Belorussians, Jews, or other non-Estonian nationalities. Further, to focus the study on intra- and intergenerational change, the new survey interviewed the oldest child of each respondent, (Estonian and non-Estonian) from the class of 1966. We refer to the two generations as the "parenthal" and the "youth" cohorts or generations. The new survey was carried out between February and July of 1991. By good fortune, the extended field date of our survey during this part of 1991 was not punctuated by events that led to sharp swings in the public mood or political activity. Estonia was in hiatus between a period of strong emotions and heightened political mobilization, and a period when the new Estonian democracy could begin to stand on its own legs. Even on the vital issue of support for Estonian independence there was no shift in sentiment among our respondents throughout the months that the survey was in the field. The principal aim of the survey was to study both continuity and change in the social and political life in Estonia between the late 1970s and 1991. The survey had two foci: (1) individual life course changes relating to education, marriage, childbearing, labor-force participation, and career advancement; and (2) intra- and intergenerational change in social and political attitudes. This overview make us of only part of the data collected from the survey, focusing especially on the three-year period leading up to 1991. Although perestroika began in the Soviet Union in 1985, the first real changes in Estonian society connected with perestroika began to take place in 1988. After that time, political events in Estonia developed very rapidly, and one of the aims our survey was to study behavior, attitudes, and values during this crucial period. Furthermore, our analysis was focused primarily on the data which we collected concerning the parental or older generation rather than focusing on the younger generation or on intergenerational change. This perspective was dictated by our desire to provide an accurate picture of people's lives during the critical years of Estonia's transition away from a totalitarian society. As part of our study, we investigated several aspects of the life situations and the perceptions of the respondents during Estonia's transition to independence. Generational and Ethnic Differences The 1991survey had both longitudinal and cross-sectional aspects. Of the ethnic Estonians who were intervied in the earlier rounds of the ELS, approximately 1.400 were reintervied in 199l. In 1979, the respondents were about age 31; in 1991 they were about age 43. The control group of non-Estonians from the same cohort was comprised of 519 respondents. Together these two groups of respondents comprise what we shall call the "older cohort" or the "parenthal cohort". If these respondents had children who were older than 16, then they were also interviewed. Children were identified and located through their parents. In this way 778 Estonian and 219 non-Estonian "children" were interviewed. This cohort is referred to as the "children's" or "youth" cohort in our study. This sampling strategy was the result of a conscious decision about how to take advantage of the existing data from 1979. Because they completed general secondary school, the respondents from the Estonian older generation represented the upper 50 percent in educational attainment of their cohort. Seventy-five percent of them received further education after completion of general secondary school, by either completing higher education or at least enrolling higher education for a while. Non-Estonians in the study were chosen from those who had an extended opportunity to adapt life in Estonia. All had lived in Estonia for at least twenty-five years. Most were born in Estonia. Because of both their educational levels and their long-term residence in Estonia, the Estonian and Russian respondents in the older cohort were more similar in background than the "average" Estonian and Russian of the same cohort. While we were conscious of the ways in which the samples were not representative of the larger population of Estonia, we sought to maximize the similarity across nationalities and generations through our sample design. This sample design also allowed us to take full advantage of the fact that older cohort of Estonians had been interviewed before, and hence, we had substantial amount of information about their values and behaviors at an earlier stage of their life course. Questionnaires Each questionnaire had an Estonian and Russian language variant. In the design of parental survey, we adapted a number of items from the previous stages of ELS as well as the All-Union Longitudinal Survey that had been conducted in 1982-83 and in 1987-88 in fifteen regions of former Soviet Union, including Estonia. A substantial part of survey involved collecting "objective" data on the life course: education, work, migration, fertility, family, and so forth. This focus on the collection of data concerning the life course was consistent with the primary focus of the previous waves of the ELS and is also the main subject of several articles included in these theme issues. We replicated also from previous waves of the ELS batteries of questions on political and social activities as well as on core values such as people's occupational and life goals. However, because one of our most important goals was to access how people's behaviors and attitudes had shifted over time in response to the dramatic changes that were taking place in Estonia , the 1991 survey instrument contained a much richer set of questions than previous stages of the ELS about political activity as well as several sets of questions about political attitudes and beliefs. Furthermore, in 1991 was possible for the survey to address aspects of people's values and life histories that could not be addressed under the prevailing conditions in Estonia when the previous stages of the ELS were conducted. Foremost among these were questions about people's attitudes toward Estonian independence, attitudes concerning different ethnic groups, experiences and memories of repression, involvement in private-sector work, and inequality of the sexes. These topics were not included by political reasons. Some other topics, like participation in "unorganized" mass political activity and attitudes toward marketization and engagement in the private economy were suggested by the dramatic changes that occurred in Estonia during the last few years. In the construction of questionnaire we followed the strategy to place our 1991 study into comparative context. We did not want to include questions in the survey that would become outdated by the time the field period was completed. The translation and informal pretesting of drafts of the instruments took place primarily in Estonia (but also partly in the United States) during the summer and fall of 1990. We did not want to include questions or phrases that could prove to be antiquated or irrelevant, or that might become catch phrases of particular political movements or parties and hence suggest to the respondents that we had a particular political agenda in asking our questions. Therefore, in framing questions about political issues, we focused on issues that had been the subject of active political discussion or legislative debate. We tried to be concrete. We attempted to stay away from extremely topical formulas and expressions that might send a false signal of our own political orientations to the respondents. Our overall goal was to keep the questionnaire for the parents to an average of less than two hours in length and for the children less than one hour. In actuality, on average, the interviews with the parents lasted 118 minutes, while the interviews with the children lasted 80 minutes. Field Work Language The original questionnaires were drafted in English during a series of collaborative meetings between the Estonian, Russian, and American researchers. Complete translations were made by a native Russian speaker and native Estonian speaker. Interviewers The early stages of ELS had been conducted in a self-administered paper-and-pencil format. The administration of the first survey in 1966 was perhaps simplest of all because it was done in the classrooms. For the 1979 survey, however, it was decided that the structure of instruments had became too complex to rely on a self-administered format and interviewers were used. The interviewers were hired through academic networks and both Estonians and Russians were used. Pretest and Field Stage The pretest was conducted in January 1991 using approximately 25 interviews and contained mainly new questions. Attempts to contact respondents began in October 1990 when letters were sent to those people who had been selected for the sample. Addresses were obtained from the Estonian address bureau in Tallinn. The potential respondents were required to return postcard indicating if they were willing to participate in the study. Interviewers were assigned lists of respondents, names, addresses, and if available, telephone numbers. Dates and Location of Interviews The field work started in February and was largely finished by the middle of June. Fifty-five percent of the interviews were conducted in the homes of the respondents, while 27 percent were conducted in their places of work. Another 6 percent were conducted in various public places (such as cafes), while 12 percent were conducted in special rooms at our facilities in Tallinn and Tartu. We had virtually no problems with the local authorities during the fieldwork. We were very fortunate that the survey began after tense moments of January 1991, and that the worst of those events had occurred elsewhere (for example, killings in Lithuania and Latvia). In addition, we finished the fieldwork before the aborted coup in August 1991. While residents of Estonia were certainly concerned about what was happening to their Baltic neighbors, in general the situation in Estonia was calm. Completion Rates The Estonian parent sample was derived from a list based on respondents who participated in the original 1966 wave of the survey. The list included the names of 1.714 persons, of whom 1.464 had been not only in 1966 but also in 1979, and 250 had been interviewed in 1966 but not in 1979. All told, 1.402 interviews were completed with the 1.714 persons in the Estonian parents sample for a crude completion rate 82 percent. In the 1991 survey, the non-Estonian parent sample was drawn from lists of students enrolled in Russian-language schools during the period between 1965 and 1967 (not all the schools that were contacted agreed to provide their enrollment lists). Starting with the sample of 708 people, 519 ended as completed interviews, for a crude completion range of 73 percent. Refusal rates were very low.