DATA SOURCE = RAND AMERICAN TEACHER PANEL (ATP) SURVEY
Designed to survey the same educators at regular intervals over time, RAND’s American Teacher Panel (ATP) survey launched in 2014 and has been administered multiple times yearly since. Panel members complete three or four web-based surveys every year, and panelists receive a small monetary incentive after completing each survey.
THE OVERALL ATP SAMPLE
To create the ATP, RAND researchers first sampled 2,300 public schools from the 2010/11 and 2011/12 federal Common Core of Data (managed by the U.S. Department of Education’s National Center for Education Statistics), and stratified that sample on five variables: grade span (primary, middle, high and combined); school size; poverty status (proportion receiving free/reduced-price lunch); population density; and geographic region. Then, using probability-sampling methods, two regular full-time teachers were selected from each sampled school to join the ATP, and these teachers were stratified by experience level (novice and experienced) and subject area (three strata at the elementary level, and five strata at the secondary level). Individuals who change schools remain on the panel; new members are added periodically so the panel remains representative of teachers over time. In 2015, the full ATP sample included 3,540 total teachers across the U.S., and surveys were completed by 65 percent of them.
OCTOBER 2016 CALIFORNIA ATP SAMPLE, RESPONSE RATE & MARGIN OF ERROR
A total of 523 California teachers were included in the ATP sample for the October 2016 survey administration; 281 responded, for a response rate of 54 percent. Weighting, which accounts for differential sampling and non-response, was used to produce results representative of the state sample. In turn, these survey results have an approximate margin of error of ±6.5 percent.