Data » What's Available » Metadata Summary


Correction to G1074A A3.R EDUCATION - COMBINED

2000 HRS Core (Final, Version 1.0)

10/10/2002

For your convenience, because much of Section A was asked only of new respondents, past wave data were merged for previously interviewed respondents for a few variables. Among these variables, which have an "A" extension and appear in the codebook following the associated variable, is G1074A A3.R EDUCATION - COMBINED. Data from preload variable G229 PR229.PREV WAVE R EDUCATION YEARS was combined with data from G1074 A3.R EDUCATION to create G1074A.

However because of incomplete preload information, 691 AHEAD respondents previously interviewed in 1995 have an incorrect value for G1074A. All AHEAD respondents who, in 1995, had a value of 8 or 9 for D649 A3.R EDUCATION had, in error, a value of 0 for the 2000 preload variable G229.

We have re-released the file H00A_R.DA with a corrected value for G1074A. If you have previously downloaded the data, we suggest that you download the corrected data and replace G1074A with the new variable. Alternatively, if you have the education data from AHEAD 1995, you may correct the values of G1074A as described below. In either case, 463 cases will change from 0 to 8; 228 cases will change from 0 to 9. Frequencies for G1074A before and after correction are listed below.

Sample SAS code for making the correction using the values of D649 from 1995 is show below. You will want to provide equivalent statements if you are using another software package.




  Distribution before correction                Distribution after to correction

      A3.R EDUCATION - COMBINED			      A3.R EDUCATION - COMBINED

                         Cumulative		                         Cumulative
  G1074A    Frequency     Frequency		  G1074A    Frequency     Frequency
  ャャャャャャャャャャャャャャャャ		  ャャャャャャャャャャャャャャャャ
       0         901           901		       0         210           210
       1          58           959		       1          58           268
       2          86          1045		       2          86           354
       3         222          1267		       3         222           576
       4         188          1455		       4         188           764
       5         229          1684		       5         229           993
       6         379          2063		       6         379          1372
       7         367          2430		       7         367          1739
       8         671          3101		       8        1134          2873
       9         519          3620		       9         747          3620
      10        1105          4725		      10        1105          4725
      11         994          5719		      11         994          5719
      12        6566         12285		      12        6566         12285
      13        1299         13584		      13        1299         13584
      14        1711         15295		      14        1711         15295
      15         656         15951		      15         656         15951
      16        1757         17708		      16        1757         17708
      17        1835         19543		      17        1835         19543
      97          36         19579		      97          36         19579
      98           1         19580		      98           1         19580


/*
SAS correction statements for G1074A A3.R EDUCATION - COMBINED
2000 HRS Core (Final, Version 1.0)
09/25/2002
*/
*******************************************************************;
*       subset data including HHID PN
	with D649A from 1995 and G1074A from 2000;
data da95;
	set in95.A95a_r;
	keep hhid pn d649a;
run;
data da00;
	set in00.H00A_R;
	keep hhid pn g1074a;
run;
*******************************************************************;
*       sort each subset by hhid pn;
proc sort data=da95;
	by hhid pn;
run;
proc sort data=da00;
	by hhid pn;
run;
*******************************************************************;
*       check frequencies prior to correction;
proc freq data=da00;
	title 'Distribution before correction';
	table g1074a / nocol norow nopercent missing;
run;
*******************************************************************;
*       merge using hhid pn;
*       output only cases in 2000 dataset;
*       reassign values for cases with value of 0 for g1074a and
*       value of 8 or 9 for d649a;
data da;
	merge da95 da00(in=a);
	by hhid pn;
	if a;
	if g1074a eq 0 and d649a in (8 9) then g1074a=d649a;
run;
*******************************************************************;
*       check frequencies after correction;
proc freq data=da;
	title 'Distribution after to correction';
	table g1074a / nocol norow nopercent missing;
run;
*******************************************************************;


Copyright © 2014 The Regents of the University of Michigan