ipating by telephone or online. Telephone interviews were conducted from Harris's
telephone research center in Youngstown, Ohio
Harris drew a random sample of 7500 current public school teachers through-
out the United States from a list compiled by Market Data Retrieval of Westport,
Conn. Sample sites for completed interviews were set for state, based on statistics of
public school teachers in each state published by the U.S. National Center of
Education Statistics.
Selected teachers were sent an alert letter containing a brief description of the
survey process and some background information on the Harris organization in order
to maximize response rates. Teachers were given the option of participating by tele-
phone or online and provided with a Harris toll-free telephone number and with a
website URL and unique password. Each selected teacher who did not participate
online was contacted at his or her school by telephone and invited to participate in
the survey. If the teacher was not available, a message was left, including a toll-free
number to allow a return call.
Before being asked to complete the actual interview, each teacher was screened
to ensure that he or she is currently teaching at least part-time in a public school and
currently teaches in grades 7 through 12. If the interview was being conducted by
telephone, and the respondent passed the screen, an appointment was made to com-
plete the interview at a time convenient for the teacher. If the interview was being
conducted online, the respondent completed the interview at that time.
Weighting of Teacher Data
The sample was not weighted since its distribution by the key variable of race already
reflected the actual proportions in the population.
Parent Sample
Telephone Survey
The survey included a total of 256 interviews with a nationally representative sample
of parents of public school students in grades K through 12. Interviews averaged 15
minutes in length and were conducted between April 18, 2000 and April 30, 2000.
Interviews were conducted from Harris's telephone research center in
Binghamton, NY.
Eligible households were identified from a random national sample. The Harris
Interactive Inc. National Telephone Sample is based on a methodology designed to
produce representative samples of persons in telephone households in the 48 conti-
nental United States. The Harris National Telephone Sample makes use of random
digit selection procedures that assure equal representation of persons in households
"listed" in telephone directories, as well as persons in households which are "unlist-
ed" in telephone directories.
1
The Harris National Telephone Sample is selected by
a three-stage, stratified sampling process. The ultimate result of this process is a set
of sample selections (telephone numbers). The sample is explicitly designed to assure
the proper representation of households in central city, suburban, and rural areas
within each of the 48 continental states.
120
1
Some households are "unlisted" as the result of a request for an unlisted phone number by the tele-
phone subscriber. Other households are "unlisted" in the published directory because the telephone num-
ber was assigned after the publication date of the directory. Samples that are restricted to directory listed
numbers only may contain serious sample biases because of the exclusion of various types of unlisted
households.
