TABLE 2-3 Selected Unweighted Response Rates for the SESTAT
|
|
Components (in percent) |
Conditional Response Rate Unconditional Response Rate |
||
|
Year |
NSCG (Reported) |
NSCG (Computed)a |
NSRCG (Reported) |
SDR (Reported) |
|
1993 |
78 |
78 |
84 |
87 |
|
1995 |
95 |
74 |
83 |
85b |
|
1997 |
94 |
70 |
81 |
84 |
|
1999 |
91 |
63 |
79 |
82 |
|
SOURCE: Data from Westat (2002a: Table 1), amended according to personal communication from SRS staff. Notes: See text for discussion. For background information also see design and methodology report on the National Science Foundation SESTAT website (http://srsstats.sbe.nsf.gov/techinfo.html). aThe computed rates for 1995 and later are approximate. The committee acknowledges that these rates are based on assumptions that the committee has not been able to verify. bIn 1995 a subsample of mail nonrespondents was selected for computer-assisted telephone interview (CATI) follow-up. The unweighted response rate using the total sample (including those subsampled out) as the base is 77 percent. The weighted response rate was 85 percent, as shown in the table. |
||||