Skip main content and go to side content

Where am I?

Retaining Teacher Talent

Research Study Methodology

Telephone Survey

Data were collected by phone and online between April 16 and June 22, 2009. In order to capture the views of younger teachers as well as the perspectives of teachers overall, the survey included an oversample of 241 teachers aged 32 and under; 642 interviews were conducted with teachers age 33 or older. The response rate for the survey was 25 percent.
The margin of error for the report is plus or minus 4.4 percent. However, it is higher when comparing subgroups or question items that weren’t asked of all respondents.

Sample Design

The sample was provided by Market Data Retrieval (MDR). A disproportionately stratified sample along with screening was used to over-represent teachers under age 32. The sample was drawn from MDR’s master list of classroom teachers who teach in public schools in the continental United States and then stratified by school type and tenure to ensure sufficient young and older teachers across school types for the analysis.


The questionnaire was designed by Public Agenda, and all interpretation of the data was done by Public Agenda. As in all surveys, question order effects and other non-sampling sources of error can sometimes affect results. Steps were taken to minimize these, including pre-testing the survey instrument and randomizing the order in which some question and answer categories were read. The data were collected by Princeton Survey Research Associates International.

Weighting Procedures

Survey data were weighted to (1) adjust for the fact that not all survey respondents were selected with the same probability and (2) account for gaps in coverage and non-response biases in the survey frame. Weights were applied to balance gender, age, race/ethnicity, education, region, year started in the teaching profession, and type of school. The parameters for sex, age, race/ethnicity, and education came from the National Center for Education Statistics Schools and Staffing Survey. The region parameter was derived from the Census Bureau’s 2006 Annual Social and Economic Supplement Survey data. Parameters for years in profession and type of school came from MDR sample frame counts. The overall design effect of the weighting procedure is 1.82.


In the first report in the series, The View from Generation Y, we compared the views of Generation Y teachers (those aged 32 and under) to those teachers from older generations.

For the second report in the series, Teaching for a Living, we conducted a “cluster analysis,” or a series of statistical tests, to group data based on response patterns. Rather than using predetermined definitions, the technique explored how the teachers group naturally. Nearly every variable of the study was included in the analysis, utilizing all questions that were asked of the total sample and also the teacher's demographic characteristics.

The result of those tests is three groups: the Idealists (23 percent of survey respondents), the Contented (37 percent), and the Disheartened (40 percent). Each of the three groups has a distinctive set of attitudes and concerns that shape how they approach their careers and the education profession. We believe this is even more significant than the traditional separations by age, school type, race, or income, although the segments often coalesced along demographic lines.

For the third report in the series, Convergence and Contradictions in Teachers’ Perceptions of Policy Reform Ideas, we used factor analysis and a priori notions of what it means to be “effective” to divide teachers’ responses into two groups: “self-perceived effective teachers” (n = 292) and “all other teachers” (n = 598).

Self-perceived effective teachers answered four items in the following ways:

  • They reported that the subject matter test scores of their students increased a lot from the beginning of the year (versus increased somewhat, did not increase, or decreased).
  • They chose the statement, “Good teachers can lead all students to learn, even those from poor families or [those who] have uninvolved parents” as being closer to their view than the statement, “It is too hard for even good teachers to overcome these barriers.”
  • They were either very or somewhat confident that most of their students will learn the skills and knowledge they were supposed to by the end of the year.
  • They reported being very or somewhat confident that they could turn around their hardest to reach students by the end of the year.
Focus Groups

Focus groups allow for an in-depth, qualitative exploration of the dynamics underlying teacher’s attitudes toward complex issues. Insights from participants in these focus groups were important to the survey design, and actual quotes were drawn from the focus group to give voice to attitudes captured statistically through the surveys. All focus groups were moderated by Public Agenda or Learning Point Associates staff.

Four focus groups were conducted with a randomly selected group of young teachers in Long Beach, California; Madison, Wisconsin; New Orleans, Louisiana; and Washington, D.C. An additional four focus groups were conducted with teachers who had been recognized for their effectiveness by their local school district (for example, through performance bonuses) in Raleigh, North Carolina; Denver, Colorado; Houston, Texas; and Chicago, Illinois.




Go back to main content | Go back to main navigation

Go back to main content | Go back to main navigation

Go back to main content | Go back to main navigation