From its establishment in 1947 to today, the Roper Center has played a key role in the development of the field of public opinion research.

1947-1957 – Birth of the Center

“A tabulating machine which does ‘everything but scramble eggs for breakfast” arrived last month to enable the school to make use of data from a series of public opinion polls donated to Williams in 1943 by Elmo Roper, head of the Roper Poll survey group.” The Williams Record, April 28, 1948

The machine was key. Without it, the 250,000 interviews Elmo Roper had conducted for Fortune magazine and for private industry and given to Williams College were simply stacks of punchcards in heavy boxes. The cutting-edge technology, which had been used by the military in the war to manage personnel information, allowed the cards to speak.

Few universities possessed such a device, but then again, Williams was the first to have the type of raw data this technology was meant to decipher. Elmo Roper, former jewelry salesman turned market researcher and public opinion pioneer, first gave Williams College materials from his collection shortly after WWII. He chose the school as a repository because one of his sons, who was killed in the war, had been a student there.

Roper believed that the polls he conducted for Fortune magazine and other clients might form a useful resource for historians of the future. With donations of both his data and his cash, Roper intended to create an archive to lay the foundation for the newly developing field of polling research. At the time, Roper was an innovator, but others soon followed suit. The Gallup Poll archived its materials at Princeton, the Minnesota Poll at the University of Minnesota.

Unfortunately, early use of these materials was limited. York Lucci, writing in a report for the Ford Foundation in 1957, noted that the equipment and personnel required by survey research methods was the equivalent of laboratory work in the physical sciences, but hardly any universities had appropriate resources. Lucci saw the secondary analysis of completed polls as a way to overcome the financial barriers posed by those conducting original research: the legions of interviewers, the technological necessities. The obvious way forward for social science was to gather together the works that had already been done and put them at the disposal of those with the skill to make use of them.

abacus-sm

This abacus was given to Elmo Roper on-air by CBS just after the failure of the polls to accurately predict the outcome of the 1948 presidential election. Roper gave it to the Roper Center for safe-keeping, to remind the industry that, though the polls may falter from time to time, those in the field should always strive to improve upon the art and science of survey research.

Lucci went on to postulate the need for a center which would collect, index, publicize in academic journals, and make available to scholars the polling data available from commercial and academic polling units in the U.S. and abroad. He hoped that such a center would allow for new ways of conducting research, such as compiling data from multiple surveys, and would be a treasure trove for future historians and current academic investigators alike.

Elmo Roper had come to the same conclusions. In May 1957, what had been the Roper Collection at Williams officially became the Roper Public Opinion Research Center. The board of advisers was replaced with a board of directors stocked with some of the biggest names in public opinion and market research, including George Gallup, Clyde Hart of NORC, and Sam Stouffer of Harvard. Although Roper’s materials formed the core of the collection, additional survey organizations began to contribute data, which was then made available to outside researchers. All interested academics were encouraged to visit the Center to do research, request basic analysis to be run for them by the Center, or even request duplicates of the punch cards.

The letters came rolling in. During the five years before the Center opened, there had been on average 15 petitions yearly to utilize the Roper data. In the first two months of the new Center, 100 requests were received. And, led by the leading lights of polling research – Roper, Gallup and Archibald Crossley – increasing numbers of pollsters began providing data. By December of 1957, nineteen American organizations – including AIPO (Gallup), Belden Associates (Texas Poll), Crossley, Field Research (California Poll), Minnesota Poll, NORC, and the Office of Public Opinion Research at Princeton, were contributors – and plans were underway to approach firms in Europe, Latin America and Australia as well. The first director of the Center was Philip Hastings, a professor of psychology at Williams who not only led the collection efforts for the US data collections, but was also responsible for archiving surveys from several European polling organizations.

1958-1976 – Expansion

The expansion of data holdings increased at a rapid rate. By 1963, there were about 2,800 studies in the collection – by 1967, 7,000. Regular reports offered updates on newly available data to the academic libraries and other organizations that formed the membership base for the Center. Academic papers began to appear in journals outlining the Roper Center holdings and the significance of the data for various fields of inquiry, including population data, sociology of religion, and trend studies. Technological changes had come to the field. Punch cards had widely been replaced by magnetic tapes, which made accessing and analyzing the data easier. Users no longer had to physically visit the Center or have boxes of punched cards shipped to them. Magnetic tapes containing copies of pertinent dataset files along with paper copies of documentation could be mailed to researchers for them to run analyses locally.

1977- Today – Continued Growth and Innovation

A structural reorganization in 1977 led the Center to the University of Connecticut, its home for the next three decades. Williams maintained a role in managing special projects at the Center, while Yale managed user services until 1985. By the time of the relocation, the collection included 9,000 studies from 75 countries. Technologically, the Center remained on the cutting edge of social science data analysis. By 1980 Gammell and Grandon wrote in Data Bases in the Humanities and the Social Sciences of the powerful new ability for Center staff to machine-search keywords on text for parts of the collection, and the exciting plans to provide the users of future with data access via a long-distance network.

The executive director at the time was Political Scientist Everett Carll Ladd. As director for more than twenty years, Ladd established the Center as a resource for the understanding the public in the aggregate. With his extensive publications offering analysis of thousands of polls, across subgroups and countries, Ladd became an authoritative public figure in the interpretation of polls. In 1986, he furthered his vision of the Center as a major institution for the understanding of public opinion by establishing Public Perspective, for which he served as editor. From 1989- 2003, the magazine offered vital information about public opinion, polling methodology, and issues in the field of survey research and greatly increased Roper Center’s impact in the world of public opinion. Also in 1986, POLL (Public Opinion Location Library) became a reality: a database where users could search, over a telephone line, most of the Roper Center’s US holdings by cross-indexing topics, keywords, organizations and dates. In March, 1989, Information Today awarded POLL the first Buddie award for Best Unknown Database, noting “it is hard to imagine a database that would find wider use by business, government, academia and the public.”

Since the turn of the century, the Roper Center has again turned its emphasis to expanding its archival holdings and improving access with creative new user tools. The Center has continued to innovate, offering RoperExpress, instant downloads of datasets, in 2005; RoperExplorer, an online statistical analysis tool, in 2011, and iPOLL +, instant crosstabs on major demographic measures, in 2011. Through its long, strong relationships with the active researchers in the polling community, the Center has built its collection to include over 22,000 datasets from more than 100 countries, and over 600,000 questions in iPOLL. Over 250 membership organizations utilize the collection, which has served as resources for countless books, articles and media stories.