February 5, 2015

Investigative Journalists and Digital Security

Methodology

The analysis in this report is based on an online survey conducted from December 3 – 28, 2014, among a representative sample of 671 journalists who are members of Investigative Reporters and Editors Inc. (IRE). The survey was written and analyzed by Pew Research Center staff, and field work was conducted by Princeton Survey Research Associates International (PSRAI). The report was produced in association with Columbia University’s Tow Center for Digital Journalism.

The universe of potential respondents for this study was defined as investigative journalists in the United States who are members of Investigative Reporters and Editors (IRE). Founded in 1975, IRE is a nonprofit association “dedicated to improving the quality of investigative reporting,” according to its website. For the purpose of this study, the total member list was filtered, and members who were not based in the United States, or who did not identify themselves as practicing journalists (educators or students, for instance) were excluded from further sampling. After an initial list of domestic, practicing journalists was pulled (a total of 2,734 records) from the IRE database, eleven records were removed because they had no email address, were based outside of the United States or contained generic contact information for an unnamed respondent. No sampling was done among the remaining 2,723 records. All domestic, practicing investigative journalists were asked to participate in the study. The overall response rate was 24.6%.

On December 1, 2014, each journalist with a postal address was sent a postcard in the mail explaining the purpose of the study and encouraging the journalist’s participation. Following the postcards, all requests for participation were sent by email. On December 3, all members were sent email invitations with individualized survey links to access the online study [N=2,723]. Email reminders were sent to all non-responders on December 10 [N=2,438] and December 17 [N=2,247]. Journalists also had the option of returning a print version of the survey through the mail. The survey was conducted in English.

The 671 investigative journalists includes a mix of reporters, producers, editors, data specialists, photojournalists and others, and comprises a wide range of coverage areas — a mix of national, local and international, including (but not limited to) government and politics; crime and law enforcement; national security; foreign affairs; education; the environment; business; and science and technology. According to the survey, a majority of respondents (88%) say they work for a news organization, as opposed to freelancing. One-in-five (21%) cover primarily national news, and 63% cover mainly local news; the remainder cover international news (3%) or report no geographic focus to their work (12%). The sample includes a mix of new, mid-career and veteran journalists and is evenly divided between those under age 40 (49%) and those age 40 and older (49%).

To assess representativeness of these 671 journalists against the total population surveyed (2,723), a series of tests were conducted based on available, anonymized membership data on both groups (a different source of member information than the survey itself). First, the groups were sorted by employment status: Among respondents, 92% listed a specific organization as their employer, while the remainder were either listed as “independent” or had a blank field (among the total population, 93% listed a specific organization as their employer). Second, the state of residence for each member was tallied and sorted. Out of the top 10 states listed for each group, nine were the same for both respondents and non-respondents (California, New York, the District of Columbia, Florida, Texas, Illinois, Virginia, Massachusetts and Pennsylvania). Third, members were sorted by region of the country and density of the city they listed — the differences between those of respondents and non-respondents were not significant. Fourth, members’ employers were classified into categories and assessed for frequency. Among the group of respondents, print-based organizations were the most frequently listed, at 47%, followed by broadcast (24%), ‘Other’ (19%) and wire services (3%). Independent journalists or blank fields accounted for 8%. Among the total population, print was also the most frequently listed, at 43%, followed by broadcast (29%), ‘Other’ (17%), and wire services (4%). Independent journalists or blank fields accounted for 7%. Finally, aggregate data on one demographic characteristic — sex — was compared between the group of respondents and the entire IRE membership. Among all IRE members who completed this field (including international and student/academic members), 56% are male and 44% are female. Among respondents (which do not include international or student/academics), 55% are male and 45% are female.

The margin of error for all respondents [N = 671] is plus or minus 3.8 percentage points. Sample sizes and sampling errors for subgroups are available upon request.