Employees as Subjects in Applied Research
There are many scenarios under which investigators may wish to ask employees to serve as subjects. Three typical scenarios are:
- Consultancy: At an organization’s request, the investigator collects and analyzes data. The investigator’s findings are “fed back” to the organization in some form, but are not made publicly available. This scenario does not qualify as “research” under the federal regulations, and so does not require review and approval the University Committee on Activities Involving Human Subjects (UCAIHS).
- Pure research: Prompted by his/her desire to answer a scholarly research question, the investigator finds an organization that is willing to permit him/her to collect data from the employees of that organization. Findings are disseminated to a scholarly audience. This scenario is clearly research, and as such requires review and approval by UCAIHS before being initiated.
- Applied research: At an organization’s request, the investigator collects and analyzes data. The investigator’s findings are “fed back” to the organization in some form, and are also disseminated publicly (typically, to a scholarly audience).
This scenario is also research, but has raised several issues during UCAIHS reviews. This following discussion is meant to identify some of those questions, and supply guidance about what UCAIHS has generally found to be acceptable practice. However, each case is unique, and this document is not intended to guarantee any particular Committee decision.
Given that employees are often good sources of information about organizations, and in light of the asymmetrical employer-employee relationship, applied organizational research raises several ethical issues. The potential for employee coercion to participate in research is perhaps the most important of these issues, since employees often feel that they must perform at the behest of their employers. Thus, investigators must take steps to minimize employee coercion, while maintaining strong study validity.
Identifying and contacting potential subjects
Investigators may request that employers or managers identify potential employee subjects. It is permissible for employers to provide employee contact information at work. Researchers may contact employees via in-house mail. Such mailings may include the following:
- A cover letter from the employer that briefly explains the purpose of the research. The stated purpose should be truthful and transparent. For instance, if the purpose of the research is to collect information that might plausibly lead to the termination of employees, the letter should so state.
The letter should clarify that (1) participation is voluntary (that employees are free to decline to participate; that employer won’t know who participates in the research; that there will be no retribution for non-participants); (2) participation is permitted (for instance, that it is OK to meet with the researchers during business hours).
- A letter from the researchers reiterating the purpose of the research and voluntariness, and explaining the procedures for enrollment.
Potential subjects may be contacted directly, or may be invited to contact the investigators. Examples of each are given below:
- Direct contact: Employee is contacted in a phone call during work hours. In the course of that conversation, study purpose and voluntariness are reiterated and the invitation to participate made. If subject chooses to participate, a mutually convenient time for interview is scheduled.
- Invitation to contact: The abovementioned letter from the researcher includes the investigator’s contact telephone number. Interested employees may call to arrange an interview.
Note: given the busy pace of work life in many organizations, invitation to contact may yield an insufficient response rate, which could invalidate the research. While direct contact is somewhat more intrusive, it is not inherently coercive. Moreover, it is likely to yield a higher response rate, and thus more valid study findings.
To the extent possible, employees should be interviewed or surveyed in a fashion that does not reveal to the employer whether or not they are participating in the study.
Reporting to employers and/or scholarly audiences
In reporting research findings, to the extent possible:
- information about participants should be presented in a de-identified fashion (eg. “one employee said ….”, rather than “John Doe said ….”). Special attention should be given to masking the identity of participants, when there are relatively small numbers of subjects who might have contributed the information.
- information about non-participation should be presented in a de-identified fashion (eg. “90% of the employees participated … “ rather than “none of the employees in X division agreed to participate”). Special attention should be given to masking the identity of non-participants, when there are relatively small numbers of potential subject who might have declined to participate.