MethodBack to Table of Contents | Back to Findings
- Study focused on United States institutions with a Carnegie classification of "Research Extensive", with the New York Times used as a proxy for high-prestige national stature news visbility. Institutions with multiple campuses are grouped together as a single entry.
- The complete contents of the New York Times East Coast Late Edition from 1945 to 2005 (18 million total documents) was keyword searched for any mention of a United States research university.
Access to the New York Times East Coast Late Edition is provided through Proquest Corporation's Historical Newspapers database, which runs through 2005, encompassing more than 18 million documents. (Post-2005 content is available through a separate Proquest database and does not use the same content tags to allow seamless querying.) The 2004 Department of Education's Integrated Postsecondary Education Data System (IPEDS) Institutional Characteristics survey was used to compile a list of all universities and colleges active in that year. Comparing a large research university with an associate's college would not yield meaningful results, so institutions were grouped by their Carnegie Classification cohorts. A random selection of institutions in each category suggested that those in the "Research Extensive" classification yielded the greatest New York Times coverage and so analysis was restricted to those institutions. Coverage was selected beginning with 1945 in order to capture the entire post-World War II era.
The IPEDS survey includes separate entries for each campus of multi-campus institutions. The University of California, for example, has eleven separate entries for its campuses and central administration office. News coverage in the New York Times, however, rarely addresses a campus by name, such as "University of Illinois at Urbana-Champaign." Patents and NSF grants are similarly listed at the institutional, not campus, level and many institutions only report budget information in aggregate. Campuses are therefore grouped together under the institution's primary name and the aggregate total of news volume and institutional characteristics are considered.
The New York Times is the paper of record for the United States and its high-profile national audience is highly sought after by research universities. While major state papers like the Chicago Tribune and Los Angeles Times are read nationally, none has the prestige or reach of the New York Times. Attempting to compare individual state papers' coverage of their local institutions is also problematic in that each paper may have different selection biases, making interstate comparisons difficult. Focusing exclusively on the New York Times enables direct comparison of institutional coverage nationally.
A simple keyword search is used to return all news articles mentioning an institution's name anywhere in the title or body text. Wedding announcements often mention the university the bride and groom graduated from, as do obituaries. These are not directly related to an institution and often have more to do with the location or status of the individuals in the announcement than the institution itself and so are dropped from this analysis. Only Times content classified as a news article is examined. The search does not attempt to measure the centrality of the school to the article's topic. For example, in 2003 then-Attorney General Eliot Spitzer of New York gave a speech at Adelphi University where he announced he would be suing a unit of Dow Chemical Corporation. (Spitzer, online) The only relation to Adelphi University was the setting for Spitzer's speech, but such a mention would be counted under this scheme, since it places the institution's name in the national media.
To maximize precision, only the full name of each institution is used. Articles with titles such as "Prof. C.S. Draper of M.I.T. Gets Sylvanus Albert Reed Prize" (Draper, 1946) that use an institution's abbreviation usually spell out its full name in the beginning of the article and use the abbreviation only in the title or in subsequent mentions. Abbreviations are also more prone to false matches due to digitization errors. Relying only on the full name of each institution will reduce the overall volume count for some, especially ivy league schools like Harvard University and Columbia University, which are often referred to simply as "Harvard" or "Columbia." However, a search for "Columbia" alone will return matches for the country and movie studio of the same name. The 1976 The New York Times Manual of Style and Usage (Jordan) includes entries for only a small number of higher education institutions, but suggests they be spelled out in full, with abbreviations restricted only to direct quotations or subsequent references. That is not to say this guideline is always followed, but it is difficult to train a machine to reliably differentiate between a reference to "Princeton" the city and Princeton the university.
The IPEDS Institutional Characteristics file contains an Institution Name Alias field where institutions can self-report common alternative names, including abbreviations. There were 103 institutions reporting a value for this field in 2004, with the majority being abbreviations. Some institutions, such as the University of Illinois, include a large number of entries not directly related to their institution, such as the name of the city they are located in, surrounding cities, and misspellings:
- Jordan, Lewis. (1976). The New York Times Manual of Style and Usage. New York City: The New York Times Book Company.
- Spitzer Says He'll Sue Dow Chemical Unit. (2003, April 3). The New York Times. http://www.nytimes.com/2003/04/03/business/spitzer-says-he-ll-sue-dow-chemical-unit.html
- Prof. C.S. Draper of M.I.T. Gets Sylvanus Albert Reed Prize. (1946, January 28). The New York Times. p. 21.
- U.S. Dept. of Education, National Center for Education Statistics. Integrated Postsecondary Education Data System (IPEDS): Earned Degrees. Washington, DC: U.S. Dept. of Education, National Center for Education Statistics.