Most academics would agree that data should guide policy. In 2020, we formed a grassroots team of early career researchers (ECRs) from the University of Melbourne and Monash University to develop a survey to gather evidence on our colleagues’ workplace culture and mental health issues, with the end goal of bringing about positive change at our universities. This initiative, which collected responses from hundreds of ECRs, has thrown a spotlight on the taxing work environment academics face and, critically, has led to university policy changes, from streamlining promotion processes to increasing the length of contracts. Our findings have also been covered extensively in the press and we’re preparing the survey for a national roll-out at institutions across Australia.
Launching such an endeavour and broaching sensitive topics, such as bullying, academic misconduct and sexual harassment, brings practical and ethical challenges but is a necessary step towards fostering progress. Here are seven tips and lessons we learned that may help other ECRs design surveys at their own university.
- Three key tips to help balance research and teaching duties
- Let’s get professional: advice for new university teachers
- Five strategies to generate a sense of satisfaction in your academic work
1. Have your survey run by ECRs, for ECRs
Having ECRs retain full control of how survey results are disseminated – regardless of what displeasing information may come to light – will aid you in gaining trust from your participants. Likewise, obtaining ethics approval to protect anonymity and to safeguard ownership of the acquired data will allow you to protect the identity of your participants. Set and state a clear plan early that the report will be available to all participants and the community.
2. Use validated instruments from the academic literature
Thousands of assessment tools are available to examine mental health and workplace behaviours. A select few, however, have sound psychometric properties and are widely used in practice. Consulting researchers in the fields of psychology and public health will assist in identifying the most suitable instruments for the subject you aim to investigate. A valid tool will also enable you to compare your findings with other surveys and to track your cohort longitudinally. For example, our team used the Generalised Anxiety Disorder Assessment (GAD-7) to screen for general anxiety disorder prevalence and to contrast our results with research conducted in other populations.
3. Acknowledge that your survey can’t include everything and choose your priorities early
Breaking news: us ECRs are overworked and time-poor. Remember that your participants are doing you a favour by completing your survey and be mindful of their busy schedules. We chose to design a survey that wouldn’t take more than half an hour to complete – a decision that meant several pertinent topics couldn’t make the final cut. Settle on a set number of fundamental domains early on and avoid having meetings dissolve into “wouldn’t it be interesting if we looked at…” sessions. Your participants will thank you.
4. Constructive collaboration with leadership increases your chances of having meaningful impact
There is a certain truth to the saying “you can catch more flies with honey than with vinegar”. Working with, instead of against, leadership is more likely to produce better outcomes. Our group was fortunate to receive support from senior leadership at both our universities to carry out our survey, which not only allowed us to use university resources to spur participation and dissemination, but also enabled us to regularly provide input on how our discoveries could inform new policy.
5. Have a diverse team with varying experiences and backgrounds
By having our team include ECRs with different experiences, we tapped into an exceptionally diverse talent pool featuring multiple disciplines and backgrounds. This breadth led us to probe topics that may have otherwise been neglected, for example, inquiring about visa status, disability status and caring responsibilities. By the same token, this functional diversity will better place your group to succeed in recruiting a group of respondents that better reflects the ECR community at your university.
6. Define roles and set realistic deadlines
There is more to running a survey than you might think: data storage, publicity, statistical analysis and drafting a report are just a few of the associated tasks. Designating roles that complement your team’s skill set will allow you to avoid confusion over who is responsible for specific duties and increase accountability. Give team members more time than they think they need to meet deadlines and underscore the importance of dedicated alone time to complete their tasks. During dissemination, devise a divide-and-conquer strategy to present and communicate the survey findings to all relevant avenues.
7. Empower participants by asking them to suggest solutions
If your intention is to pinpoint the struggles your ECR community is facing, ask your survey participants to propose solutions to improve their workplace. Apart from having participants rank potential changes in policy they would like to see implemented, we offered participants the opportunity to provide free-form text answers. Some highlights included increasing ECR representation on decision-making bodies and promoting project management skills training.
Developing an effective ECR survey requires a considerable amount of time and effort. We hope the tips outlined here set your team on the path towards maximising your survey’s impact.
Trevor Steward is senior research fellow; Hui-Fern Koay, postdoctoral research fellow; Kelly Kirkland, postdoctoral research fellow; George Taiaroa, postdoctoral research fellow; and Claudia Marck, senior research fellow, all at the University of Melbourne. Darshini Ayton is senior research fellow at Monash University.
If you found this interesting and want advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.