What can the National Child Development Study (NCDS) and the 1970 British Birth Cohort Study (BCS70) tell us about the individual’s acquisition and use of education and training?
What can the National Child Development Study (NCDS) and the 1970 British Birth Cohort Study (BCS70) tell us about the individual’s acquisition and use of education and training?
Key takeaways
(file:///C:\Users\scott\Downloads\BCS%20DUMP\mdwp.pdf)
Bibliography: Makepeace, G., Dolton, P., 2001. What can the National Child Development Study (NCDS) and the 1970 British Birth Cohort Study (BCS70) tell us about the individual’s acquisition and use of education and training? CLS Cohort Studies.
Authors:: G Makepeace, P Dolton
Collections:: UCL BCS Dump
First-page: 3
Abstract
At the end of 1999, the QCA commissioned the Centre for Longitudinal Studies to undertake a review of research that uses NCDS and BCS70. The review specification focused on the role of education and training in the career progression of individuals. It entailed a survey of past and present research users of NCDS and BCS70 and the production of a bibliography of published and unpublished work that uses these data. The Appendix describes the survey and the resulting references have been incorporated into the Bibliography. The review begins with a short description of the main characteristics of the cohort studies and the advantages and disadvantages of working with such data. It then examines the five research areas that the QCA identified as of particular interest to them. We shall describe them in the order they appear in this report.
Citations
content: "@makepeaceWhatCanNational2001" -file:@makepeaceWhatCanNational2001
Reading notes
Imported on 2024-05-07 21:38
⭐ Important
- & The main advantage of these studies for economic analysis is that there is a lot of detailed information on a lot of individuals. More importantly, there is data on the same individuals over time at different stages in their life. The relatively large number of observations gives robust estimates of the parameters in statistical models. It facilitates the analysis of subgroups so, for instance, separate models can be estimated for men and women and greater levels of disaggregation such as by region can be considered. Special groups such as apprentices7 can be studied. (p. 3)
- & Although not perfect, the data especially for NCDS has been carefully collected in some detail and many important variables are included in the data. NCDS, for example, collects diary information that allows work histories to be constructed. This means work experience will be estimated more accurately compared with data sources that rely on approximations8. One problem is that the data collection has not been unified so, for example, BCS70 does not collect dated work-life histories at age 26. Some variables have attracted widespread attention. NCDS provides measures of reading, mathematics and general ability at 7, 11 and 16, obtained from tests that are independent of educational qualifications. (p. 3)
- Super important to the conversation on Age/Period/Cohort:
- & The data is by definition representative of a cohort. The sample members have similar histories so certain things such as government policy that change over time and affect economic outcomes are the same for all observations. In that sense, any analysis can automatically control for these effects. (p. 5)
- & Each cohort may have experienced very different conditions from one another and from those existing today. For instance, the NCDS cohort were educated under an ‘GCE/CSE’ regime before comprehensive schooling was universal. There was a relatively active youth market operating before the decline of apprenticeships in the 1980s and with no government sponsored youth training schemes. Between the dates when the NCDS members and BCS70 members reached the age of 16, the fraction of 17 and 18 year olds participating in full time education more than doubled. Female participation in education and the labour market also increased significantly over this period. (p. 5)
- & The attrition problem is magnified by item non-response. Item non-response refers to the fact that there are typically missing values for key variables in a piece of analysis. Thus some respondents will not have reported their wage while others may not taken an ability test. Most researchers deal with this problem in a pragmatic way by dropping observations with missing values on any variable. Sometimes an attempt is made to retrieve observations by including a dummy variable for missing values (p. 6)
- & The BCS70 survey at 26 did not collect detailed work history data although the 10% sample at age 21 did. (p. 14)
- & The picture for BCS70 is less clear because there is no month-by-month report of the respondent’s activities. Joshi and Paci (1997) comment “A summary of job history once in the labour market was pieced together from answers to various questions on spells in various employment states, and length of time in current job. (p. 14)
- & Schoon et al (2001) presents a direct comparison of the work histories of individuals between the ages of 16 and 21 based on the work history information in NCDS and the 10% sample of BCS70 members at age 21 (p. 15)
- & Work history information is collected at the time of the survey and is retrospective. For example, data for the 17-year period from 1974 to 1991 is collected in 1991. There is a potentially serious problem if respondents cannot recall information accurately over time. Sometimes the same information is collected at different points in time56. Dolton and Taylor (p. 15)
- This is the argument I use for the BCS data where some of the data isn't identical when item response is from both sweeps. Dominance to earlier sweep approach is best and justified again here.:
- & (1999) note that work experience can be computed for the period 1974 to 1981 using either NCDS IV or NCDS V. They argue that the NCDS IV data should give a more accurate representation of the work histories over the period 1974 to 1981, because the individuals were nearer the period at issue at the time of that survey (p. 16)