The Project to Document the History of Physicists in Industry: Some Notes on Methodology
The Project to Document the History of Physicists in Industry ends this December, and so far this year we’ve completed the last of the site visits and interviews at industrial labs—at Raytheon in January and Ford in June—and focused on analyzing the 132 interviews that we’ve conducted along with other information that we’ve collected. When we planned the study, we decided that individual interviews with physicists, R&D managers, and information professionals (e.g., technical librarians, archivists, and records managers) who work at 15 of the 27 largest employers of physicists in industry would be the best way to capture the experience and perspectives of the participants with as much richness and context as possible.
Business in general has frequently been described as one of the least documented sectors in American society, and sources on the work of corporate physicists are especially rare. So our purpose has been to learn as much as we can about the extent to which these records do exist; how companies treat correspondence (including e-mail), lab notebooks, and other documentary materials of scientists today; the effect of the computer revolution on records keeping; and other information that will help us make informed recommendations on appropriate strategies to identify and preserve vital elements of this virtually unknown history. In addition, because the study provides the opportunity to meet and interview a cross section of physicists at 15 of the country’s largest high-tech firms, we have also included questions about career patterns in industry, how R&D is funded and structured, and other questions that will give a better understanding of the nature of industrial physics over the past 25 years.
The kind of interview-based qualitative research that has been the backbone of our study is probably less familiar to most people than quantitative research, which typically relies on statistical surveys and questionnaires of large samples of people. Surveys and questionnaires are excellent instruments for identifying trends and variances distributed over large numbers of people, and the findings are usually presented in terms of percentages of participants who fit into certain predefined categories. Interviews, on the other hand, provide the stories behind the percentages by allowing study participants to describe the context within which they develop their careers, make decisions, and keep records.
Broadly, qualitative research describes inquiry into perspectives, processes, and themes. Quantitative research, in contrast, describes inquiry into trends and variances from those trends. Because these two complementary research approaches espouse fundamentally different goals, their underlying assumptions, data collection methods, analysis procedures, validity checks, and outcomes differ quite markedly from each other. These differences do not necessarily mean that qualitative and quantitative research methods are incompatible with each other, or that one tradition delivers “truer” results than the other. Rather, the nature of the underlying research question–the impetus of the research project–should dictate which methodology is most germane.
We prepared for the study by conducting an initial mail survey of information professionals at approximately 40 high-tech companies. The results were inconclusive, however, and the qualitative approach that we’ve employed in the study has allowed us to start out with broad research questions like “What kinds of company policies and procedures affect the maintenance and completeness of lab-notebooks?” We suspected that each company’s policies would differ from each other, and these broad questions have allowed study participants to provide thorough and individualized answers.
Fortunately and coincidentally in 2006, AIP’s Statistical Research Center conducted an Industrial Membership Survey for the American Physical Society. The survey instrument was sent to 2,700 APS members in the U.S. with private company addresses, and 1,200 people responded. Some of the questions, including those dealing with application of physics knowledge on the job, favorite information sources, and networking styles, overlap with some of the themes of our interviews.
We described some of our preliminary findings in this newsletter in the Fall 2006 issue, and we are now analyzing the thousands of pages of interview transcripts in Nvivo, a qualitative analysis software that allows us to create a hierarchically arranged catalog of the major and minor concepts and themes that showed up across our data set. NVivo’s query tools then allow us to filter and navigate these concepts according to company, industry sector, or interviewee characteristics. We will also compare the results of our findings with the APS survey responses, all of which should provide a solid basis for our recommendations and report. Our more interesting findings include the extent to which companies are grappling with problems in the conversion from a paper to an electronics based records system.
While no company thinks it has resolved all the problems, several have developed innovative programs which others may find useful. Project Staff will present papers on our analysis-to-date at the annual meeting of the Society for the History of Technology in mid-October and will complete and circulate a draft report for comment by the end of the year.