Online Survey
The survey questionnaire was self-administered online by means of the Internet to
1761 parents of public school students in grades K through 12 who have access to the
Internet. Interviews averaged 15 minutes in length and were conducted April 17,
2000 and April 26, 2000.
Sample was drawn from the Harris Poll Online database of approximately 6.5
million households who are registered as participants in Harris's online database.
Email addresses for households in the database are obtained from participation in the
following sources: Harris Poll Online registration, the Harris/Excite Poll, Harris
Poll Online banner advertisements, Excite and Netscape product registrations,
Harris telephone research, media recruitment, and sweepstakes sponsored by
Matchlogic and its subsidiaries (e.g., DeliverE, Preferences.com, etc.). Invitations for
this study were emailed to a selected sample of the database identified as being 25 to
59 years old and residing in the U.S.A.
Weighting of the Parent Data
The completed interviews were weighted to the U.S. Census Bureau's latest popula-
tion parameters on education, gender, race or ethnicity and age. This weighting
adjusted these key variables where necessary to their actual proportions in the popu-
lation. In addition, data collected online were weighted to key behavioral and attitu-
dinal variables to align it with the data collected using the Harris National Telephone
sample and to represent two-thirds of the total interviews.
Telephone Interviewing Procedures (Teachers and Parents)
Interviewing for this study was conducted by Harris's professional interviewing staff
and was continuously quality monitored by the supervisory staff. Through direct
supervision of the interviewing staff and continuous monitoring of the interviews, a
uniformity of responses was achieved that could not have been obtained by other
interviewing methods.
The Harris computer assisted telephone interviewing system (CATI) permits
on-line data entry and editing of telephone interviews. Questionnaires are pro-
grammed into the system with the following checks:
1. Question and response series
2. Skip patterns
3. Question rotation
4. Range checks
5. Mathematical checks
6. Consistency checks
7. Special edit procedures
The CATI system reduces clerical error by eliminating the need for keypunch-
ing, since interviewers enter the respondents' answers directly into a computer dur-
ing the interview itself. For questions with pre-coded responses, the system only
121
