Table A-2.

Design effects (DEFF) and root design effects (DEFT) for selected High School and Beyond surveys and subsamples: 1984 and 1986

Subsample characteristic	Follow-up survey of 1980 sophomores in 1984		Follow-up survey of 1980 seniors in 1984		Follow-up survey of 1980 sophomores in 1986		Follow-up survey of 1980 seniors in 1986
	DEFF	DEFT	DEFF	DEFT	DEFF	DEFT	DEFF	DEFT
Total sample	2.40	1.54	2.87	1.69	2.19	1.47	2.28	1.50
Sex
Male	—	†	—	†	2.07	1.43	2.13	1.45
Female	—	†	—	†	2.06	1.43	2.26	1.50
Race/ethnicity
White and other	2.06	1.42	2.09	1.44	1.92	1.38	1.70	1.30
Black	2.22	1.47	2.26	1.50	2.19	1.47	2.40	1.54
Hispanic	3.15	1.73	3.72	1.92	3.11	1.76	4.06	2.01
Socioeconomic status   composite (SES)1
Low	1.91	1.37	2.28	1.50	1.83	1.35	2.31	1.51
Middle	1.95	1.39	1.81	1.34	2.06	1.42	2.02	1.42
High	2.05	1.42	1.93	1.38	1.92	1.38	1.71	1.30

— Not available. † Not applicable. 1 The SES index is a composite of five equally weighted measures: father's education, mother's education, family income, father's occupation, and presence of certain items in the respondent's household. NOTE: The average design effect for the 1980 sophomore cohort first follow-up (1982) survey is 3.6 (1.89) and the average design effect for the 1980 senior first follow-up (1982) survey is 2.6 (1.62). Race categories exclude persons of Hispanic ehtnicity. SOURCE: U.S. Department of Education, National Center for Education Statistics, High School and Beyond Study of 1980 Sophomores (HS&B-So:80/84 and HS&B-So:80/86); and High School and Beyond Study of 1980 Seniors (HS&B-Sr:80/84 and HS&B-SR:80/86).