‘Speaking Freely’ FAQ: FIRE answers Frequently Asked Questions about new survey

Today, FIRE released “Speaking Freely: What Students Think about Expression at American Colleges,” a groundbreaking survey of student attitudes toward free speech.

Here are the answers to what we anticipate will be the most Frequently Asked Questions about the report.

What is FIRE?

The Foundation for Individual Rights in Education, or FIRE, is a nonpartisan, nonprofit educational foundation based in Philadelphia, PA. FIRE’s mission is to defend and sustain individual rights at America’s colleges and universities. These rights include freedom of speech, legal equality, due process, religious liberty, and sanctity of conscience — the essential qualities of individual liberty and dignity. FIRE’s core mission is to protect the unprotected and to educate the public and communities of concerned Americans about the threats to these rights on our campuses and about the means to preserve them.

Why did FIRE pursue this survey?

A previous survey by the Knight Foundation and Gallup found that college students support First Amendment protections of free speech and expression in theory. We were interested in gaining a deeper understanding of what this support meant when students encounter issues surrounding free expression in their everyday lives. To do so, we tried to answer the following questions:

• Are students comfortable expressing themselves on campus? Are they expressing their opinions and ideas on campus?
  
• How are students reacting to the speech of their peers?
• What are students’ attitudes toward guest speakers, protest, and hate speech?
• How supportive do students feel their campus is toward students?

By publishing the results of this survey, we hope to contribute to a national conversation about free expression at American colleges and universities.

Who administered the survey?

FIRE contracted with YouGov (California), a nonpartisan polling and research firm. YouGov has polled for the New York Times and the Economist. The Pew Research Center finds that YouGov outperforms its competitors when it comes to providing accurate survey results.

How many people were interviewed?

YouGov used an online survey to interview a total of 1,395 two- and four- year undergraduate students at American colleges and universities. Once YouGov completed the interviews, they provided us with a final dataset of responses from 1,250 students.

When was the survey conducted?

Students responded to the survey between May 25, 2017 and June 8, 2017.

What is the margin of error for the survey?

Tabulations from the final dataset have an estimated margin of error of plus or minus 3.1 percentage points at the 95 percent confidence level. Tabulations taken from subgroups of the sample have a greater margin of error.

The margin of error and confidence level in a survey are a measurement of how accurate the tabulations are and how confident someone can be in the tabulations. For example, we report that 48 percent of students support First Amendment protections of hate speech. Our margin of error means that our result may differ up to 3.1 percentage points in either direction, so that in the general population between 44.9 percent and 51.1 percent of students support First Amendment protections of hate speech. Our confidence level means that if our survey was conducted 100 times, we would find the same results for this statistic 95 of those times.

What is a non-probability sample and why did FIRE use one for its survey?

We decided to pursue a non-probability methodology after considering the cost, speed, and data quality that we needed for this project. Like the New York Times and the Economist, we decided to contract with YouGov, the most accurate non-probability polling company.

Non-probability sampling is a new methodology in the survey field, and has strengths and weaknesses — just like more traditional probability sampling. The main weakness of non-probability sampling is that, unlike in a probability sample, not every individual in the population of interest has the same likelihood of being sampled for response. In other words, there is concern that data collected with this method does not accurately reflect the views of the population polled. To make up for this weakness, a matching and weighting process is undertaken so that the non-probability sample reflects the population of interest as closely as possible. We looked at the tabulations for both the weighted and unweighted data, and the results were extremely similar.

How was your sample weighted?

Our poll was administered online to a sample of YouGov’s panel of respondents. After students responded to the survey, YouGov matched the responses down to our final sample using a sampling frame from the 2013 American Community Survey. Weights for each response were calculated by YouGov based on the respondent’s gender, race, and age.

What other surveys have been done on this subject?

Very few surveys have specifically looked at the attitudes of American college students toward free expression.

• The Higher Education Research Institute (HERI) at UCLA has published statistics on the attitudes of students toward expression in many of their The American Freshman: National Norms publications.
• In 2016, the Knight Foundation and Gallup released a survey that found college students are confident in their right to free expression, but that they feel the campus environment may not be open to all ideas.
• The William F. Buckley, Jr. Program at Yale University put out national polls looking into campus censorship in 2015 and 2016. In 2017 they distributed a similar survey only to Yale students.
• A current project by John Villasenor, a professor at UCLA and a nonresident senior fellow at the Brookings Institution, looks at students’ attitudes toward hate speech and guest speakers. The study has been partially replicated by the Economist and YouGov. Although Villasenor’s survey questions have some similarities to ours, many of the results diverge substantially. The separate times at which our respective respondents were surveyed may account for some of these differences. Villasenor’s respondents were surveyed in late August 2017, immediately after the violence in Charlottesville, whereas our survey was distributed in May 2017.
• Finally, the 2011 book The Still Divided Academy, a follow-up to 1976’s The Divided Academy, assessed the attitudes of administrators, faculty, and students toward the academy. The authors include statistics on students’ feelings about censorship, which are consistent with the findings in our survey.

Other surveys have looked at the attitudes of American adults toward free expression.

• The General Social Survey has asked Americans whom they would support speaking in their communities since the 1970s. This data has recently been discussed by the American Enterprise Institute, and The Washington Post’s Wonkblog and Monkey Cage.
• The Newseum has released a series of surveys on American’s knowledge of their constitutional rights in State of the First Amendment reports. Similar civic knowledge surveys have been pursued by other organizations, such as the Annenberg Public Policy Center.
• Recently, the Bucknell Institute for Public Policy released a report on free expression and due process. The Cato Institute will soon be releasing a report on adults’ attitudes toward free expression.

This is not a comprehensive list of surveys on free expression, but we hope that it is a helpful starting point for research in the area.

Is more research needed on this topic?

Although surveys have analyzed students’ knowledge of their First Amendment rights, and some of their attitudes toward expression on campus, more work should always be done. To our knowledge, no survey before ours has taken more than a cursory look at students’ attitudes toward guest speakers or how students define hate speech. These are topics ripe for discussion and more research. Additionally, replication is an important part of the research process. So that trends in public opinion over time can be analyzed, researchers must ask the same questions of similar populations over time.

How was the survey funded?

The survey and resulting report was made possible through the support of a grant from the John Templeton Foundation. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the John Templeton Foundation.

How can I access your survey’s data?

We have made the tabulations and toplines available on our website. The survey questionnaire and its specifications are found at the end of the survey report. We will make the respondent-level data public in the summer of 2018.

I’m a member of the media, who should I contact to talk about the survey?

All media requests can be emailed to media@thefire.org, or you can contact Communications Coordinator Bill Rickards at 215-717-3473.