Survey of Socioeconomic Opportunity and Achievement 1991-1994 - South Africa

Abstract
---------------------------

The survey was conducted from 1991 to 1994 by a research team headed by Donald J. Treiman, UCLA; Sylvia N. Moeno, then in the Strategic Planning Group at Eskom, Johannesburg; and Lawrence Schlemmer, then Director of the Centre for Policy Studies, School of Business, University of the Witwatersrand, Johannesburg. Data collection was undertaken by MarkData, the in-house survey unit of the Human Sciences Research Council (HSRC), Pretoria, South Africa. The survey was funded by the U.S. National Science Foundation (SES89 12677), and by several South African agencies: the Anglo- American/De Beers Chairman's Fund, the Trust for Educational Advancement in South Africa, the Human Sciences Research Council, Johannesburg Consolidated Investments, and the Union Carbide Corporation.

Because at the time of the survey, South Africa had a policy of residential segregation by race (authorized under the the "Group Areas Act", which was abolished in June, 1991), South African surveys typically were conducted by drawing separate samples from each of the “Group Areas,” that is, areas designated for residence by Whites, Asians, Coloureds, and Blacks.1 In South Africa, interviewing usually is carried out by interviewers of the same race as interviewees. To accommodate scheduling considerations, the main survey was carried out in two parts: 3,679 interviews were conducted in White, Asian, and Coloured areas in April and May 1991 (the WAC sample) and 3,689 interviews were conducted in Black areas in August and September 1991 (the Black sample). Processing of the two parts was also carried out separately. Because there proved to be substantial under-representation of males in both samples (see below), data were collected for a supplementary sample of 749 men in May-July 1992. Finally, data were collected from a special sample of 969 rural respondents in 1993-94 (see below). In all, between 1991-94 data were collected for 9,086 respondents.

Geographic coverage
---------------------------

The survey had national coverage

Analysis unit
---------------------------

Households and individuals

Universe
---------------------------

The population surveyed consisted of persons age 20 or older residing in "greater South Africa," that is, the Republic of South Africa plus the nominally independent "TVBC states": Transkei, Venda, Bophutatswana, and Ciskei.

Kind of data
---------------------------

Sample survey data

Sampling procedure
---------------------------

The population surveyed consisted of persons age 20 or older residing in "greater South Africa," that is, the Republic of South Africa plus the-at the time-nominally independent "TVBC states": Transkei, Venda, Bophutatswana, and Ceskei.2 The sample, as designed, was a complex area probability sample, in which sampling categories and the number of cases to be interviewed were determined by the researchers, households were to be selected within areas at random with probability proportionate to size, and adults within households were to be selected at random, but in such a way that the probability that males would fall into the sample was twice the probability that females would fall into the sample. The sample was designed so that, suitably weighted, it could be regarded as representative of the entire adult population of "greater South Africa." Further information on sampling is in the survey codebook provided with the data.

Mode of data collection
---------------------------

Face-to-face [f2f]

Data appraisal
---------------------------

In addition to the codebook the documentation for this survey included three reports on the field work. The first report, Social Stratification Report, provided information on the initial WAC sample but has not been located. This report contains, among other information, a copy of the English language questionnaire used for the WAC sample and responses to open-ended questions in the WAC sample. The second report, Fieldwork Report - Project: Social Stratification, provided information on the Black sample, the special samples, and the supplementary sample of males but has not been located.

摘要
---------------------------
本调查由唐纳德·J·特里曼教授领导的科研团队于1991年至1994年期间进行，团队成员包括加州大学洛杉矶分校的西尔维亚·N·莫诺，当时任约翰内斯堡 Eskom 战略规划小组成员；以及约翰内斯堡维特沃特斯兰德大学商学院政策研究中心主任劳伦斯·舍勒默。数据收集工作由南非比勒陀利亚人类科学研究中心（HSRC）的内部调查单位 MarkData 承担。调查资金由美国国家科学基金会（SES89 12677）以及多个南非机构提供，包括英美/戴比尔斯董事长基金、南非教育进步信托基金、人类科学研究中心、约翰内斯堡综合投资公司和联合碳化物公司。
由于当时南非实行基于种族的居住隔离政策（该政策由《集团地区法》授权，并于1991年6月被废除），南非的调查通常是通过从各自的“集团地区”（即专为白人、亚洲人、有色人和黑人指定的居住区）抽取单独样本来进行的。在南非，访谈通常由与受访者同种族的访谈员进行。为了适应调度考虑，主要调查分为两部分进行：1991年4月至5月，在白人、亚洲人和有色人居住区进行了3,679次访谈（WAC样本），1991年8月至9月，在黑人居住区进行了3,689次访谈（黑人样本）。两个部分的处理也是分开进行的。由于两个样本中男性的代表性均存在较大不足（详见下文），因此于1992年5月至7月收集了749名男性的补充样本。最后，1993-94年间，从969名农村受访者中收集了数据。总的来说，1991-94年间共收集了9,086名受访者的数据。
地理覆盖范围
---------------------------
调查覆盖全国。
分析单元
---------------------------
家庭和个人
总体
---------------------------
被调查的人口为居住在“大南非”的20岁及以上的人员，即南非共和国以及当时名义上独立但实质上受控的“TVBC州”：特兰斯凯、温达、博普塔茨瓦纳和齐塞凯。
数据类型
---------------------------
样本调查数据
抽样程序
---------------------------
被调查的人口为居住在“大南非”的20岁及以上的人员，即南非共和国以及当时名义上独立但实质上受控的“TVBC州”：特兰斯凯、温达、博普塔茨瓦纳和齐塞凯。设计的样本为一个复杂的多区域概率样本，其中抽样类别和访谈案例的数量由研究人员确定，家庭将在区域内随机选择，且选择的概率与规模成比例，家庭内的成年人将随机选择，但选择的方式使得男性进入样本的概率是女性进入样本概率的两倍。该样本设计得如此，在适当加权后，可以被视为“大南非”整个成年人口的代表性样本。有关抽样的更多信息，请参阅随数据提供的调查手册。
数据收集方式
---------------------------
面对面访谈
数据评估
---------------------------
除了代码手册外，本调查的文档还包括三个关于实地工作的报告。第一份报告《社会分层报告》提供了关于初始WAC样本的信息，但尚未找到。该报告包含其他信息，如WAC样本使用的英语语言问卷副本以及WAC样本中开放式问题的回答。第二份报告《实地工作报告 - 项目：社会分层》提供了关于黑人样本、特殊样本和男性补充样本的信息，但尚未找到。