The survey was conducted through Survey Monkey, a Web-based survey instrument. Prior to beginning the survey, two preliminary e-mails were sent in order to confirm that the most appropriate person within the organization had been contacted and to apprise respondents of the purpose of the survey. As many as eight follow-up e-mails were sent to organizations that had not responded over a period of one month, from April 25, 2006, to May 26, 2006.
The sample for this survey was compiled from the databases of the American Enterprise Institute, the Heritage Foundation, the Mackinac Center for Public Policy, the State Policy Network, All Children Matter and the State Policy Network 2005 annual conference. Although most of the groups contacted were known by the author and the Mackinac Center to be directly involved in school choice or general education reform, an effort was also made to include groups that were planning to become involved or whose primary purpose is a policy area like tax reform, which is directly affected by school choice policy.
The sample of 473 organizations contained groups from every part of the political and ideological spectrum. A substantial portion of the respondents were self-identified "free-market" organizations. Most major state and national free-market think tanks, most major state and national school choice issue organizations, and most major state and national fiscal, religious and family issue organizations were contacted. The total survey response rate was just under 50 percent (233 organizations), with 37 percent completing question 75. Additional "dropoff " occurred thereafter, with 35 percent (167 organizations) answering through the end of the demographic section.
As can be seen in the demographic questions at the end of the questionnaire, the organizations that responded to this survey reflect some of the diversity of the school choice movement, and the survey provides a wide-ranging look at organizational and professional opinion. As is common with this kind of research instrument, the survey collected the opinions of only a sample of all state and national organizations involved in the movement. Thus, as with most surveys, response bias poses a threat to the validity of the results. Respondent diversity and the large percentage of state organizations responding are encouraging in this regard, but the only way to dispel all possibility of response bias is to succeed in surveying the vast majority of organizations involved in school choice. A similar survey may thus be attempted in the future.