Archived Information

Endnotes

¹A Carnegie Unit is astandardized measure of coursetaking. One Carnegie Unit is earned for every course that meets for five 50-55 minute periods per week for an entire school year. Thus, a one semester-course would earn one-half unit.

²Achievement scores used in these and the following analyses in this report are the Item Response Theory-scaled number correct (IRT Number Right) in mathematics, science, and reading.

³ Curricular program type was defined by categories of the variable F2RSPFLG, which uses the NAEP:87 criterion for defining program type using Carnegie units for courses taken. The Academic category in Table 2 includes course patterns labeled "Rigid Academic", "Academic", "Rigid Academic and Vocational", and "Academic and Vocational". The Vocational category included students who met the criteria for Vocational Program, but not the criteria for Academic or Rigid Academic Program. Rigid Academic and Vocational and Academic and Vocational were included in the Academic category for analyses conducted for this report. The General category included students who did not meet the criteria for either the Academic or Vocational Programs. Further details are provided by Ingels et al. (1994), Appendix H, p. 27.

⁴ All statistical tests were calculated using the SUDAAN program. This program uses information about the survey design to produce estimates of standard errors. Survey design information relevant to the NELS:88 data are stratification, clustering, oversampling, and weighting to account for nonrespondents.

⁵ The unstandardized regression coefficient is the predicted increase or decrease in 12th-grade achievement for every unit increase in coursetaking. For example, if the unstandardized regression coefficient between math achievement and coursetaking in calculus were 1.50, the predicted difference between students who took no calculus and students who took one Carnegie Unit of calculus would be one-and-a-half achievement points.

⁶ Two-stage least squares regression was implemented by conducting two separate regression analyses. The first analysis regressed 10th-grade achievement on 9th- and 10th-grade coursetaking (instrumental variables), 11th- and 12th-grade coursetaking, and background characteristics.

⁷ I am grateful to Professor Robert Meyer of the University of Chicago for pointing out this problem in his review of this manuscript.

[References ]