E ditor's Note: This is the final article in a three-part series about the University of Chicago Biological Sciences Division Postdoc Association (BSD-PDA). Earlier articles told of the inspiration to found the association  and the process of developing, conducting, and analyzing a survey  of the UC BSD postdoc population. In this article, we learn how survey results can be used to advocate for change at the administrative level.
We had completed the first survey of postdocs at the University of Chicago (UC) and felt a sense of accomplishment. As it turned out, the work had actually just begun. The feeling, probably familiar to many of you, was that of a submitted article coming back from review as "accepted pending revisions." Although pleased that your experiments have been evaluated favorably and the main bulk of the writing is done, you find yourself a bit dismayed that more effort is required to refine the data presentation and interpretation. You wonder if the additional effort will be worthwhile.
We had taken a lot of time to design our survey and a lot more to compile and analyze the results. The Postdoc Association (PDA) membership and some UC administrators, department chairs, and deans, served as reviewers and helped us refine our discussion. The Biological Sciences Division (BSD) Dean of Graduate Affairs acted as our editor, facilitating contact between us and most of our reviewers. These efforts initiated a necessary dialogue.
The survey means different things to different people. Our goal to improve the "postdoctoral training environment" evolved. The complexity of the decision-making process in the UC BSD became apparent as we observed different perspectives and goals at different levels. For example, we noticed that departments had a relatively high degree of autonomy, but administrators in the dean's office wielded power of policy. Even with a supportive interim dean, our long-term efforts to encourage a broad review of postdoctoral policies were significantly delayed by the search for a permanent dean.
Interpreting Survey Data and Finding a "Hook"
Our primary goal for the survey was to seek a mandate: to align the postdoctoral association's efforts with the priorities of our membership. The survey yielded interesting unanticipated findings. We found, for example, that within a sample of 26% of the 255 postdocs in our division, stipend and salary ranges were large at each experience level. More surprising, there were many postdocs reporting compensation well below National Institutes of Health (NIH)/National Research Service Awards (NRSA) recommended minimums. Of the many important issues that came out of our survey report, salary and stipend levels became our primary focus. Similar to a research paper, quantifiable survey data are more resistant to dismissive interpretations than more subjective perceptions of the "quality" of the training environment. This is how we got our foot in the door to discuss postdoctoral policies with UC administrators.
Building Consensus and Learning
Facilitating change requires a community of postdocs, a network of allies, and a continuing invitation to participate in policy development. The survey results focused our attention on two areas: providing career development opportunities and representing postdocs to the administration. With this direction, we developed an executive summary  in which we proposed specific action points to improve the postdoctoral training environment.
As we presented our survey results and proposed action points to department chairs, administrators, an associate provost, and the new dean, we continued to develop and hone our message. We have come to appreciate the advocacy process is an inherently idiosyncratic process: successful advocacy could play out quite differently in other environments or at other times.
In developing our community of advocates, we decided that focusing on compensation was simplistic and in addition could be misinterpreted as a chant for a postdoc union. We did not want to alienate administrators who would otherwise be helpful and supportive, so we refocused our message to advocate for the improvement of the overall training environment in a manner that would include but not be limited to efforts to provide equitable compensation.
Building consensus required inclusion, and we worked with the administration to form relevant academic committees with postdoc representation that could address postdoc issues. Each meeting presented an opportunity for us to learn more about the policy-making process. Senior administrators and other postdocs expressed broad ideas about the postdoctoral training environment. We all agreed that sound postdoc policies make the institution more competitive. One department chair thought it important to have policies that go the extra mile to enable postdocs who demonstrate success at grant proposal or fellowship writing to continue on an upward career trajectory. He suggested developing a policy that would mandate health insurance coverage regardless of funding source, so that postdocs receiving private foundation fellowships are not penalized. Beyond the salary issue, this chair was also willing to push for and track annual meetings between each postdoc and his or her mentor.
The Importance of "Negative Results"
We found that these meetings provided a window into a complicated process where others--for different reasons--had been working on similar issues for many years. There always seemed to be more to learn. The question became not whether change should occur, but where to start. We developed formal and informal contacts with postdocs, faculty, and administrators in the physical and social sciences divisions. These conversations brought forth the profound differences in the culture of each division's training environment. We also had hints that the "maximum at the minimum" compensation policy common in the life sciences, which uses the NIH/NRSA minimums as a basis, may be restricting more competitive upward compensation pressures not found in the physical sciences. Although these interactions were cordial, we found little common ground for forming a cross-division alliance.
A related important "negative result" came after meetings with officials in the provost's office. We wanted to introduce our association survey results to the UC administration. Because the BSD dean's office was in transition, we met with an associate provost to explore a university-wide (i.e., cross-division) approach to a review of postdoctoral policies. She felt that postdoc issues were important but eventually decided not to support university-wide policies because the situations for postdocs in the physical, social, and biological sciences divisions are so different. Instead, she argued that postdoc policies were the responsibility of each division. Initially we were disappointed, but we found that knowing this was incredibly helpful: We could focus our energies on creating change within our division.
We have come to realize that the major impediment to broad change in postdoctoral policies in our division was the lengthy search for and transition to a new dean. The delay may have actually been for the best. We had time to develop a compelling interpretation of our survey data and a consensus for change based on many different perspectives among our "reviewers" and our community. Our survey also served as a motivator for us: We had hard data to fall back on when doubt or delay would sap our energy. We are encouraged that our new dean has given us a chance to be heard, even before his full-time arrival.
Two years after the founding of BSD-PSA and a little over a year since the completion of our survey, we've learned that developing an advocacy network is a long-term project. The fruits of these efforts will likely be realized well after we have moved on from our current positions.