In 2017, we defined our population of interest to only include programs at universities that fit into all three of the following categories: 1) members of the National Association of University Forest Resources Programs (NAUFPR), 2) Land Grant institutions (including 1890 schools), and 3) listed in the Times Higher Education World University Rankings for 2016 – 2017. This led to a list of 28 universities. In 2022, we included all original 28 universities and added 4 Land Grant, ranked universities that became NAUFRP members since then.
For each university in both 2017 and 2022, we went through a stepwise procedure to establish if the program had an external advisory board. First, we conducted a web search for each program, navigating to each program’s website (e.g., the Department of Forest Resources and Environmental Conservation, or the College of Forestry) and searched for an advisory board webpage. If we were able to find a webpage with names and employer affiliations for an external advisory board, we used that as the primary list of names.
If we were not able to find an advisory board online, we emailed the program head (e.g., department head, dean, etc.) for information on any external advisory boards. If the program head did not respond after a second reminder email, or confirmed the program did not have an advisory board, we moved up one administrative level if the program was not already at the college level. For example, if a department head confirmed the Forest Resources department did not have an advisory board, we would then look at the College of Environment and Natural Resources. We only pursued higher administrative levels if it seemed the advisory board would still be representative of the department. To continue our example, while a College of Environment and Natural Resources would be relevant enough for inclusion, a College of Letters and Science would be too broad and would not be included. Similar to the department level, first we searched for evidence of an external advisory board on the upper administrative level on their website. If we could not find any information, we emailed the administrative head, typically a dean of a college. We emailed the upper administrative heads once.
Once we had lists of advisory board members, either from websites or provided through emails, we created a database first with the names of the advisory board members and their affiliated organizations. We sought information on board members first and last names, employer type (described in more detail below), gender, and ethnicity. Information was first drawn from advisory board websites or lists provided through emails. Missing information was filled in using a Google search. Where relevant, each member’s name was entered in the Google search window, along with their affiliated organization and the state of the university. This often yielded “Meet Our Team” pages from the employer’s website, interviews and features in local newspapers, and publicly available LinkedIn pages.
We were purposeful in performing the data recording and collection in a specific order (beginning by recording all information given in the advisory board list before supplementing it through a Google search) for the sake of consistency. For example, if a list taken from the advisory board website indicated that “Sally Jones” worked for “Green Tree Timber Co.,” but a Google search showed that she had recently moved to “Sturdy Timber, Inc.,” her affiliation was still recorded on the spreadsheet as “Green Tree Timber Co.”
The employer/affiliate categories were modified from the categories used by Sample et al. in their 2015 report and included university faculty and extension, natural resources businesses, non-natural resources businesses, natural resources consultant, federal agencies, state agencies, local government, and NGOs (Sample et al., 2015). We used employer websites, and particularly the “About” and “Mission” pages to help us categorize these organizations. The categorization was intuitive, since it was not difficult, upon navigating through each website, to determine that an organization was, for example, a timber firm, putting them into the category of “natural resources business.”
We made visual observations of photos advisory board members to categorize each member by race and gender. We also relied on text to categorize individuals. For example, if a board member website bio described them as a member of a Tribal nation, or if they were identified on the advisory board page by a prefix, such as “Mr.” or “Ms.,” or if they were referred to with a gender pronoun such as “he” or “she” in any text collected through the Google search. If we were unable to find a prefix or pronoun, we relied on the free online program Gender API (https://gender-api.com/), which draws on a database of close to two million first names to provide the user with an informed guess about the name holder. We only used guesses that were rated as 80 percent accurate and above.
To protect confidentiality and right to privacy, in the deidentified dataset we replaced university names with INSTITUTIONID and respondent names with RESPONDENTID.
