Methodology

The 2017 PDK survey is based on a random, representative, 50-state sample of 1,588 adults interviewed by cell or landline telephone, in English or Spanish, in May 2017. 

The 2017 PDK poll on education was designed, managed, analyzed, and reported by Langer Research Associates of New York, N.Y., in consultation with PDK. All results described in this report were tested for statistical significance.

Langer Research Associates is a charter member of the Transparency Initiative of the American Association for Public Opinion Research. After six months, researchers will have access to the complete dataset through the Roper Center for Public Opinion Research.

The 2017 PDK poll was conducted via the SSRS Omnibus, a national, random-digit-dialed telephone survey conducted by SSRS of Glen Mills, Pa. The SSRS Omnibus consists of about 1,000 random-sample telephone interviews per week, 600 conducted via cell phones and 400 via landline phones, with a minimum of 35 interviews in Spanish. 

Calls are made each Wednesday to Sunday or Friday to Tuesday to a fully replicated, stratified, single-stage RDD sample of landline telephone households and randomly generated cell phone numbers designed to represent the adult population of the United States. Phone numbers received up to four call attempts in a five-day period.

Within each landline household, interviewers ask to speak with the youngest adult male or female at home. Cell phone interviews are conducted with the adult answering the phone.

This year, as always, PDK has taken great care to frame poll questions as objectively as possible and to share the full and unvarnished results. 

Data are weighted via a multistage process, first correcting for unequal probabilities of selection depending on the number of adults in the household and the nature of telephone service in use, then applying a poststratification adjustment to correct for systematic nonresponse using known demographic parameters. 

The sample undergoes iterative proportional fitting (“raking”) to match the most recent March Supplement of the U.S. Census Bureau’s Current Population Survey by age (by gender), education, race/ethnicity, marital status, population density, and Census region (by gender). Respondents’ telephone status (cell phone only, landline only or mixed user) is included in the rake, based on the most recent estimates from the U.S. Centers for Disease Control’s National Health Interview Survey.

Interviews for the national PDK poll were conducted across seven waves of the SSRS Omnibus, the first among a random sample of the national population, the rest to collect additional interviews of blacks, Hispanics, and parents of school-age children. In all, 1,588 adults were interviewed May 4-21, 2017, including 636 parents of school-age children, 297 black respondents, and 289 Hispanics. 

Oversampled groups were weighted to their estimated share of the population, including parents overall and by racial/ethnic group. Results of the national poll have a margin of sampling error of plus or minus 3.5 percentage points for all adults, 5 points for parents of school-age children and 5 points for parents of public school children. These calculations include each survey’s design effect due to weighting.

The Georgia PDK poll is based on telephone interviews of 633 adult state residents May 4-June 4, 2017, including 371 parents of school-age children and 364 parents of public school children. Oversampled groups were weighted to their correct share of the population. Interviews were conducted in English and Spanish, 412 via cell phones and 221 via landlines. 

The sample was drawn from state interviews across six waves of the SSRS Omnibus survey, recontact of Georgia respondents from previous Omnibus surveys (with propensity weighting to adjust for nonresponse) and state-level random-digit dialing. Results of the Georgia survey have a margin of sampling error of plus or minus 5.5 percentage points for all adults and 7.5 points for parents of school-age children and public school parents alike. These include the survey’s design effect due to weighting.

The New York PDK poll is based on telephone interviews of 628 adult state residents May 4-June 10, 2017, including 366 parents of school-age children, 354 parents of public school children, 125 black respondents, and 143 Hispanic respondents. Oversampled groups were weighted to their correct share of the population.Interviews were conducted in English and Spanish, 362 via cell phones and 266 via landlines. 

The sample was drawn from state interviews across seven waves of the SSRS Omnibus survey, recontact of New York respondents from previous Omnibus surveys (with propensity weighting to adjust for nonresponse) and state-level random-digit dialing.Results of the New York survey have a margin of sampling error of plus or minus 5.5 percentage points for all adults and 7.0 points for parents of school-age children and public school parents alike. These include the survey’s design effect due to weighting.

The SSRS Omnibus is used by a wide range of business, media, academic, and foundation clients, including researchers from more than a dozen universities; organizations such as the Kaiser Family Foundation, the National Alliance for Hispanic Health, and the Christopher and Dana Reeve Foundation; and media outlets including the New York Times, CBS News and ABC News.

Poll archive

The PDK poll archive contains all of the questions and results from past polls. The archive is accessible with a PDK membership. In March 2018, the complete dataset will also be available through the Roper Center for Public Opinion Research