Methodology Reports

A Comparison of Results from Surveys by the Pew Research Center and Google Consumer Surveys

Assessing the Representativeness of Public Opinion Surveys

The Growing Gap between Landline and Dual Frame Election Polls; Republican Vote Share Bigger in Landline-Only Surveys

A new analysis of Pew Research Center pre-election surveys conducted this year finds that support for Republican candidates was significantly higher in samples based only on landlines than in dual frame samples that combined landline and cell phone interviews. The difference in the margin among likely voters this year is about twice as large as in 2008.

Cell Phones and Election Polls: An Update

Data from Pew Research Center polling this year suggest that the landline-only bias is as large, and potentially even larger, than it was in 2008.

Assessing the Cell Phone Challenge

Fully a quarter of the U.S. adult population now relies solely on a cell phone. This paper shows that the potential for bias in telephone surveys that do not include cell only adults has grown since 2006. Of 72 questions examined on a wide range of topics - including political and social attitudes, personal and national economic ratings, foreign policy views, and attitudes toward and the adoption of a wide range of internet and communications technologies - cell phone samples made a difference of 3 percentage points or more on 29 of the questions. In 2006, on 46 questions examined none of the differences exceeded 2 percentage points.

Accurately Locating Where Wireless Respondents Live Requires More Than A Phone Number

The mobile nature of wireless phones creates a significant problem for geographic sampling and analysis. This problem is exacerbated by the fact that the wireless-only are more geographically mobile than those with landline phones.

Perils of Polling in Election ’08

Despite such challenges as a growing wireless-only population, possible racially-related response bias and greater-than-usual difficulties in forecasting turnout, pollsters' methods were evidently adequate to the task.

Pollwatch: Comparing the Polls on Spending and the Deficit

How the question is phrased has a clear impact on whether the public rates deficit reduction or stimulus spending more important.

Why Surveys of Muslim Americans Differ

Because Muslim Americans make up a very small percentage of the U.S. public, it is difficult to provide a reliable picture of their views and differences in survey design can crucially affect findings.

New Tricks for Old — and New — Dogs: Challenges and Opportunities Facing Communications Research

Pollsters and other communications researchers are finding their job ever more challenging but also more interesting, and, with the help of new techniques and data sources, even more amenable.

Understanding Likely Voters

This analysis summarizes differences in presidential horserace estimates based on likely and registered voters in 2008 and previous election cycles and describes the Pew Research Center's 2008 likely voter scale.

Calling Cell Phones In ’08 Pre-Election Polls

The addition of cell phones in Pew Research Center election surveys had at most a modest effect on estimates of candidate support in individual surveys, but when looked at in the aggregate clear patterns emerge resulting in slightly more support for Obama and slightly less for McCain.

Poll Power

Though by no means a perfect instrument, polls make it possible for more opinions, held by a broader and more representative range of citizens, to be known to the government and thus, potentially, heeded.

Cell Phones and the 2008 Vote: An Update

As shown in our election summary release, including cell phone interviews results in slightly more support for Obama and slightly less for McCain.

Cell Phones and the 2008 Vote: An Update

This is the first study documenting how the overall estimate of voter presidential preference is modestly affected by whether or not the cell phone respondents are included.

Research Roundup: Latest Findings on Cell Phones and Polling

The Pew Research Center has been studying the challenge to survey research posed by the growing number of wireless-only households. Here's a summary of its latest findings.

A ‘Brute Force’ Estimation of the Residence Rate for Undetermined Telephone Numbers in an RDD Survey

Public Opinion Quarterly 72:28-39 (2008)
This paper discusses ''e'' the proportion of unresolved telephone numbers that are, in fact, eligible for the survey.

The Impact Of “Cell-Onlys” On Public Opinion Polling

This study finds that respondents reached on cell phones hold attitudes very similar to those reached on landline telephones on key political measures such as presidential approval, Iraq policy, vote preference and party affiliation.

“Frequently Asked Questions” about Pew’s Muslim American Survey

The Facts behind the Design, Conduct and Analysis of a High-Profile Study

A recent report, ''Muslim Americans: Middle Class and Mostly Mainstream,'' attracted a great deal of attention but also raised a number of questions about the research. Here are answers to some of the most frequently asked questions about the design, conduct and analysis of this study.

How Serious Is Polling’s Cell-Only Problem?

The landline-less are different from regular telephone users in many of their opinions and their numbers are growing fast. Can survey researchers meet this challenge?

What’s Missing from National RDD Surveys? The Impact of the Growing Cell-Only Population

This study finds evidence that as the number of cell-phone-only households has continued to grow it does not create biased estimates for the population as a whole, but it does on certain variables for young adults, 25% of whom are cell-only. A revised version of this paper was also published in

Cell-Only Voters Not Very Different: Fewer Registered, More First-time Voters

Three Pew Research Center surveys of cell-only Americans this year have found that their absence from landline surveys is not creating a measurable bias in the bottom-line findings.

Are National Polls Reliable Predictors of Midterm Elections?

This study addresses how accurate the

The Cell Phone Challenge to Survey Research

While Americans who rely solely on a cell phone for telephone service differ in their demographics from land-line subscribers, a new study finds that so far the results obtained by surveys that exclude cell-only users are not significantly affected.

Polls Face Growing Resistance, But Still Representative

Even as response rates to telephone surveys continue to decline, this study finds that carefully conducted polls continue to obtain representative samples of the public and provide accurate data about the American public.

Gauging the Impact of Growing Nonresponse on Estimates from a National RDD Telephone Survey

Public Opinion Quarterly 70: 759-779 (2006)
This article provides more extensive analysis of data from the study reported in Polls Face Growing Resistance, But Still Representative.

Screening Likely Voters: A Survey Experiment

This report presents results from a study in a closely contested mayoral race in Philadelphia conducted in 1999 to help assess which questions or combination of questions from our likely voter scale were most accurate in predicting voting.

Consequences of Reducing Nonresponses in a National Telephone Survey

Public Opinion Quarterly 64:125-148 (2000)
This article provides more extensive analysis of data from the study reported in Conservative Opinions Not Underestimated, But Racial Hostility Missed.

Online Polling Offer Mixed Results

This study compares responses from an online survey to those obtained using a random digit dial telephone survey and finds significant attitudinal and demographic differences between the general public and those who participate in online polls.

Conservative Opinions Not Underestimated, But Racial Hostility Missed

Public Opinion Quarterly 64:125-148 (2000)
This article provides more extensive analysis of data from the study reported in Conservative Opinions Not Underestimated, But Racial Hostility Missed.