May 16, 1998

Possible Consequences of Non-Response for Pre-Election Surveys

Race and Reluctant Respondents

Tables

Table 1. Initial and Reluctant Respondents: A Demographic Profile
(Based on unweighted data)

Amenable Respondents Reluctant Respondents
(No refusal) (Refusal conversion)
(N=917) (N=542)
% %
Sex
Male 45 41
Female 55 59

Race
White 79 83
Black 13 9
Other 8 8

Age
Under 30 20 18
30-49 44 40
50-64 16 20
65+ 17 17

Education
College Grad. 28 29
Some College 24 25
High School Grad. 37 37
Less than H.S. Grad 11 9

Family Income
$75,000+ 10 14
$50,000-$74,999 14 17
$30,000-$49,999 26 24
$20,000-$29,999 15 12
Less than $20,000 23 20

Region
East 20 19
Midwest 25 26
South 35 37
West 20 18

Community Size
Large City 20 20
Suburb 22 25
Small City/Town 37 36
Rural Area 19 18

Employment Status
Full-Time 55 56
Part-Time 12 8
Not Employed 32 34

Party ID
Republican 26 31
Democrat 33 33
Independent 36 30

Table 2. Demographic and methodological predictors of racial attitudes
Overall opinion Views on role of
of Blacks racial discrimination
Unstd Coeff. (S.E.) Std. Coeff. Unstd Coeff. (S.E.) Std. Coeff.
Female .107 ** (.037) .10 .127 (.080) .05

Age -.000 (.001) -.01 -.003 (.002) -.04

Education .035 ** (.012) .10 .073 ** (.026) .10

Income .000 (.010) .00 .008 (.022) .01

South -.050 (.039) -.04 -.254 ** (.083) -.10

Political ideology .044 * (.020) .07 .172 ** (.043) .13

Reluctant respondent -.122 ** (.047) -.11 -.264 ** (.099) -.11

Black interviewer .055 (.039) .05 .331 ** (.084) .14

Rigorous survey -.044 (.060) -.03 .080 (.126) .02

Constant 2.854 (.113) 1.266 (.241)

Number of cases 931 878
F-statistic 5.38 *** 9.40 ***
Adjusted R-squared .04 .08

* p < or = .05
** p < or = .01

Note: Entries are unstandardized coefficients, standard errors, and standardized coefficients from least squares regression
based on unweighted data. Dependent variable in first equation is overall opinion toward Blacks. Responses range from
1 (“very unfavorable”) to 4 (“very favorable”). Dependent variable in second equation is views concerning the main
reason “many black people can’t get ahead these days” — a forced choice between “racial discrimination” or blacks
themselves are “mostly responsible for their own condition”. Responses range from 1 (believe “very strongly” that
blacks themselves are responsible) to 4 (believe “very strongly” that racial discrimination is to blame). Full question
wording is in Appendix 2.

Response Rates
The following table presents the full disposition of sampled telephone numbers for the Standard and Rigorous surveys:

Standard
Five-Day Rigorous
Non-sample numbers:
Telephone number not in service/
not working/business number/fax number 1297 1020

Households never screened:
No answer/busy/answering machine
Not available/callback 938 185

Households that refused 678 396

Households with no eligible person:
Language barrier/health problem/
No person 18 or older 221 227

Households with eligible person:
Incomplete interviews 42 33
Completed interviews 1000 1201

TOTAL TELEPHONE NUMBERS 4176 3062

Cite this publication: “Possible Consequences of Non-Response for Pre-Election Surveys.” Pew Research Center, Washington, D.C. (May 16, 1998) http://www.people-press.org/1998/05/16/possible-consequences-of-non-response-for-pre-election-surveys/, accessed on July 22, 2014.