A thought-provoking new paper by Anne Toomey and colleagues argues that "impact is not a product to be delivered, but a possibility to be cultivated". The authors suggest moving from planning to navigating, from attribution to contribution, and from evidencing impact to sharing learning. It's a useful challenge to some of the assumptions that underpin current approaches to research impact.

A new paper by Adrián A. Díaz-Faes and colleagues examines ‘network hedging’ — the practice of consulting multiple people for the same research need rather than relying on one trusted contact for each type of advice. Although the authors apply this to citations, it applies equally to impact work: asking several people to test an idea makes us more likely to see blind spots and missed opportunities, especially when some of those people work outside academia.

A useful reminder from Livia Fritz and colleagues that no matter how respectfully we try to co-produce knowledge, our work will always be shaped by power dynamics. The paper presents three frameworks for analysing and working with power in co-production. A simple practical test: ask who has influence over the process, who benefits from the outcomes, and who is still missing from the room.

A recent paper by Nico Meissner & Kishoth Navaretnarajah argues that "serendipity is a skill that can be practiced" rather than simply a matter of chance.

Inspired by their work, here are some practical ways you can cultivate serendipity and increase your potential for impact:

• Create time for conversations beyond your immediate field
• Follow up on unexpected ideas instead of dismissing them as distractions
• Build diverse networks that expose you to different perspectives
• Make space for reflection so you can connect seemingly unrelated insights
• Treat surprises, anomalies and failures as sources of learning, not just problems to solve

Wilfred Mijnhardt has combined 20 different impact evaluation frameworks used by universities and business schools internationally to create the Impact Codex — a really useful source of ideas for measurable impacts you can plan for or evaluate across a range of categories.

Empathy is often seen as a personal quality rather than a research skill. This new paper by Melissa Robson-Williams et al. argues that empathy is fundamental to research that actually makes a difference — it helps explain why some engagement builds trust and leads to lasting relationships, while other engagement remains transactional. A useful reminder that methods matter, but relationships matter more.

We should be more conservative about claims based on evidence of immediate behaviour change. This paper by Alec Brandon and colleagues shows that social comparison nudges lead to remarkably persistent behaviour change even after they are discontinued.

Research that aligned with the policies of ruling governments was more likely to achieve impacts in REF2014. We’re re-reading this paper by Katherine E. Smith as part of our work on a blog about what to do when nobody in policy is interested in your research — watch this space.

If you haven’t already read it, we highly recommend Julie Bayley’s book Creating Meaningful Impact: The Essential Guide to Developing an Impact-Literate Mindset. It provides a thoughtful and practical perspective on building impact capacity and capability across organisations.

Franziska Sörgel and colleagues have developed a useful framework for understanding and measuring the impact of citizen participation. Their recent blog provides an accessible overview of the approach and links to the underlying Evidence & Policy publication.

Policy evidence should be judged by its fitness for purpose, not by sample size. This new paper by Anna Tarrant explains why the implicit hierarchy that privileges large quantitative samples is often critical for marginalised groups, whose voices are further marginalised by traditional evidence hierarchies.

A useful new journal for anyone interested in research systems, evaluation, research culture and impact has recently launched. The Journal of Research on Research (JROR) has been developed by an international team committed to advancing the growing field of research-on-research. We’ll be keeping an eye on articles published in the journal and sharing anything particularly useful.