RDEL #23: The most popular posts of 2023.
This week, we summarize our most popular posts of the year - from psychological safety to measuring software engineering productivity.
Welcome back to Research-Driven Engineering Leadership. Each week, we pose an interesting topic in engineering leadership, and apply the latest research in the field to drive to an answer.
❤️ This is our final RDEL of 2023 (RDEL #23 on the final post of 2023!) and we’re feeling reflective, so this week we’re rounding up the most RDEL popular posts of the year. We’ve covered various topics in software engineering leadership, including productivity, well-being, and management principles. According to our readers, these were our most popular topics:
#1: What makes a great manager of software engineers?
In this RDEL, we reviewed research from Microsoft on top attributes of a great manager, as described from engineers themselves.
Key insights were that
“Maintains a positive work environment” was the highest rated attribute for great engineering managers (mean of 9.05 out of 10)
The second and third highest rated attribute were “grows talent” (8.98) and “enables autonomy” (8.91), indicating the relative importance of helping engineers develop their talent and allowing them to organize their work.
“Being technical” was rated as 14/15 with a score of 7.84 out of 10. This was corroborated in qualitative interviews, where engineers felt that managers should have enough knowledge to “facilitate discussions that will help the engineer make decisions”, but didn’t necessarily need more.
This research highlights that in the eyes of engineers, the best managers are those that create a positive environment, foster autonomy, and help their team grow.
#2: Can better documentation come at a cost for some teammates?
This edition highlighted finding from the 2023 DORA report on the impact of documentation. While the data undoubtably showed that teams with higher documentation quality saw an increase in various technical capabilities, researchers noticed that those improvements are not spread evenly across the team.
The team found that:
Better documentation also substantially improves team performance, organizational performance, and operational performance.
Respondents across all genders saw a decrease in burnout with increased documentation, but respondents who identified as underrepresented (regardless of gender) had a higher rate of burnout.
The finding was similar for documentation quality, generative culture, and team stability: as these attributes increase, burnout also increases for underrepresented folks.
The key application for managers is to invest in documentation, but to be thoughtful about the distribution of that work - particularly if it is seen as non-promotable work that more regularly falls into the hands of women and underrepresented minorities.
#3: Is there a gender disparity in the code review process?
This edition we reviewed research from Google that analyzed thousands of code reviews performed at Google to determine if the code review process had a gender disparity.
The research found that:
Women performed 17% fewer reviews than men
Women review 16.3% fewer code reviews when the reviewers are manually assigned
Women and men are equally likely to complete an assigned review (only 0.5% difference)
This research highlights that inequities exist, particularly when reviews are manually assigned. The researchers also noted that “review suggestion tools” can sometimes present algorithmic bias. A few ways to reduce the gender disparity in code reviews is to use a randomized review tool, and to regularly audit distribution of PRs to see if any teammates are being over- or underutilized.
#4: Software Engineering Productivity: How do you measure it?
In this edition, we defined software engineering productivity by looking into the SPACE framework. (PS - in a previous edition, we established the “definition” of software engineering productivity).
In 2021, researchers at Microsoft published this groundbreaking framework for measuring developer productivity. By looking at the SPACE (Satisfaction, Perception, Activity, Collaboration/Communication, and Efficiency/Flow) of processes in software engineering, teams get a much more comprehensive picture of developer productivity. This framework also captures the tension between experience, actions, outputs, and outcomes.
For teams that are looking to quantify and improve their engineering team, Microsoft researchers recommend using a mix of metrics from at least three categories of SPACE. These can be found largely through developer surveys, as well as through system data from developer tools (i.e Github).
#5: How does psychological safety impact engineering teams?
This edition of RDEL looked at psychological safety, a widely-understood foundational element of productive engineering team. Researchers studied and surveyed 110 R&D projects across various companies to create a model for how psychological safety impacts performance.
They found that:
There was a generally positive correlation between team autonomy and psychological safety. When broken down, they saw a nuanced detail -
In low exploration projects (ie concrete projects that are clearly defined), higher autonomy correlates to higher psychological safety.
In teams with high exploration (ie huge R&D projects, new product development), higher autonomy correlates to lower psychological safety.
Higher levels of psychological safety can minimize team turnover and increase project performance.
From this work, managers can take away that psychological safety improves the performance of engineering teams, and that it is important continually measure it as projects evolve. Psychological safety is also influenced by autonomy and exploration - the best teams have the right combination of well-defined projects and a high sense of autonomy.
—
As we look back on our most popular posts, our team feels a great sense of gratitude for the many readers who have read the research, learned something new, and brought back useful insights to their engineering teams. We frequently hear from our RDEL community that our analysis offers a unique lens on nuanced engineering leadership topics, and help make difficult decisions a little easier. 🎉
We will be on a brief break until after January 1st. Wishing you and your engineering orgs a restful holiday season and a wonderful New Years.
Lizzie
From the Quotient team