Methods
I collected email addresses of 2000 faculty with field-based specializations in applied and basic biological, environmental, and geophysical science using departmental websites of a non-random selection of 200 public and private research universities, undergraduate-serving institutions, and community colleges located in the United States. The prospective respondents represented a non-random but wide range of institution types, disciplines, and geography. I then randomly selected an unstratified sample of 1000 faculty for recruitment into the survey, and contacted them via email between April 15 and April 25, 2020 (contact email details are in Supporting Information). A follow-up email was sent 5 days later to non-respondents. A unique link to the online survey was also emailed to several discipline- and interest-specific mailing lists. The survey period closed May 10, 2020.
The survey was administered using the web platform SurveyMonkey (SurveyMonkey 2020). The survey process included an informed consent statement, and consisted of 22 questions, not counting an informed consent acceptance and additional communication opt-in (informed consent, complete survey questions, and response options are reported in Supporting Information). Respondents were notified that their individual responses were private, and individually-identifiable survey data were anonymized and separated from survey responses before storage and analysis. Standard psychometric principles were not used in the creation of all survey questions given the backgrounds of the prospective respondent pool, and my intent to use these data in this purely descriptive or inductive study. Five of the 22 questions interrogated the respondent’s current teaching and plans to teach courses, institutional and positional characteristics, and specific discipline. The remaining 17 questions interrogated the instructor’s typical field instruction, and their perception of impacts of the COVID-19 pandemic and teaching modality shifts on typical field instruction (Supporting Information). Respondents were informed that they could opt out at any time, and a final submission was required for their results to be recorded. One survey response was eliminated because it contained numerous extraneous and perhaps poorly-intentioned responses. The remaining surveys were used to create simple descriptive statistics, and summary figures and tables. Not all respondents answered all questions, and thus sample sizes varied among questions are reported on a question-by-question basis. Aggregated and anonymized data are reported in Supporting Information. Individual responses to free response items are not shown as some contained potentially individually-identifiable information, and informed consent was not obtained from subjects to use direct quotations of their responses. The only person with access to non-anonymized and non-aggregated data was DC Barton. The survey and data-handling protocol was approved by the Humboldt State University Institutional Review Board for the Protection of Human Subjects (IRB #19-164) in April 2020.