Attribution – Is It Really a Big Deal? / L’Attribution – faut-il en faire un cas?

David Phipps, RIR-York

Many people worry about attribution.  How much influence did we have on an outcome? Many people except York’s Knowledge Mobilization Unit, that is. This post starts some thinking that arose with the help of the Knowledge Brokers Forum… and some time by a pool.

Nombreux sont ceux qui s’inquiètent à propos de l’attribution. Quel degré d’influence avons-nous eu sur un résultat? De nombreuses personnes, mais pas au sein de l’Unité de mobilisation des connaissances de York, cependant. Ce billet présente une réflexion qui a surgi grâce au Forum des courtiers de connaissances… et au temps passé sur le bord de la piscine.

Attribution [John Mayne, CDN J. Prog. Eval. (2001) 16(1): 1-24] is the degree to which research as well as other inputs informs a decision. If an impact happens a long time after the research is completed then it is harder to attribute impact to the research study. In a networked and complex environment there are multiple inputs into any decision confounding the ability to attribute impact to a particular study. This is generally agreed to be true. Or is it?

On November 18, 2012, I began a discussion on the Knowledge Brokers Forum listserv:


“A colleague and I were discussing the issue of attribution in knowledge exchange and I thought I would read up a little about it. So I turn to my usual sources when I want to read someone else’s work instead of do it myself (!):

  • Nutley et al: Using Evidence
  • Bennet & Bennet: Knowledge Mobilization in the Social Sciences
  • Strauss et al: Knowledge Translation in Health Care

“Attribution” is not to be found in any index in these books.

A search in google or google scholar isn’t much help because the word “attribution” usually comes up as part of a Creative Commons license.

So I am wondering… is attribution really a big deal or do we talk about it without much of an evidence base (for more on this see my earlier blog Knowledge Hypocrites)?”


I received 30 comments from around the world – for detail on comments see the compilation of comments- Attribution: KBF Responses

In York University’s Knowledge Mobilization Unit we rarely see attribution as an issue. As described in a recent book chapter our work has helped inform the cooling policies for the City of Toronto, a new funding program for United Way of York Region, a sustainability program for rural businesses and a new way of delivering immigrant settlement services in York Region as examples. When we are able to demonstrate non-academic impact of our work it is often directly attributed to the collaboration we supported. When speaking of the Green Economy Centre, Valerie Ryan, CEO of Nottawasaga Futures, said, “We could not have done this without the Knowledge Mobilization Unit”.

So what’s the big deal about attribution?

I think the role of a particular actor in knowledge mobilization/research utilization is important – see table below:

Actor Does it matter Why or Why Not?
Academic research institution No In a REF world institutions will create good news stories about any impact, no matter how small, no matter how little research contributed to the impact.
Researchers Maybe Depends on how the researcher is being measured which might depend on if they are employed by a university or a think tank or a community based researcher.
Think Tank/NGO Yes Think Tanks and NGOs are often funded based on ability to demonstrate impact and the attribution of their efforts.
Funders Maybe Private donors want to understand what impact their funding made. Academic research funding councils are beginning to care about impact but they are concerned with narratives more than attribution.
Knowledge brokers No If the goal is to broker successful collaborations or access to knowledge then the impact arising from those collaborations is nice but anecdotal.
End users No They got an impact. Attributing that to research or any other input is less relevant than the impact itself.
Evaluators Yes This is what they do. They are employed by those who care about impact and attribution.

Academic institutions routinely deal with attribution in technology transfer and commercialization. When an academic institution grants a license to a patent the attribution issue is dealt with up front by agreeing on a royalty rate. The royalty rate acknowledges there will be other inputs en route to market; however, the academic institution will create a REF case study regardless of a small or large royalty rate.

But that’s not the whole story.  A KB Forum response from Larysa Lysenko connected me to an article from Annette Boaz and colleagues. Only a portion of the article discusses attribution and in that discussion were two little sentences:

“The process by which the research is done might also have an impact on policy. For example, research in collaborative projects may have an impact prior to the production of research outputs.”

This was my “Aha Moment”. As described in a recent paper, York works almost exclusively in a co-production paradigm where the decision maker partner is an active participant in the research endeavour. The decision making process for our research partners occurs continuously during and after the collaboration. Decisions may be informed during the research process, not solely after the research is concluded.

The Knowledge Brokers (No – above) in York’s institutional (No – above) Knowledge Mobilization Unit promote a co-production method in which impacts may precede outputs. Attribution is therefore not a big deal for us. This certainly doesn’t deny the importance of Attribution in other contexts and the role of methods such as contribution analysis and productive interactions to address the issue. It’s just not a big deal for us.

And I was sitting by a pool in Ft. Lauderdale reading Annette’s paper when I had my “Aha Moment” underscoring the need to get away and just think.

David showing off his ResearchImpact T Shirt in Ft. Lauderdale

David showing off his ResearchImpact T Shirt in Ft. Lauderdale

5 Responses to “Attribution – Is It Really a Big Deal? / L’Attribution – faut-il en faire un cas?”

written by behrfacts On 23 January 2013 Reply

Nice post and T shirt. I prefer the term IMD rather than impact when speaking about affecting policy outcomes, which is why I have a doc on my website about ‘Influence, Media and Dissemination’ strategies ( which includes some case studies I was involved with in national science and maths ed policy in the UK. I agree that knowledge brokers probably don’t need this as much as others, but there is no harm in self-evaluating on a regular basis to encourage continual improvement of practice.

written by Joanne Gaudet On 28 January 2013 Reply

Very interesting. Co-production definitely makes impact matter more than attribution… research that matters.

written by Rick Riopelle On 29 January 2013 Reply

I am not sure it was an Aha! moment for me, but John Mayne’s ILAC Brief 26 (2012) on Causal Factors placed the K* Concept Paper and the literature informing that synthesis into the context of a proactive process guide which we have termed a ‘developmental nested causal factors operations model’ to frame the objectives, specific aims, and activity logic sequences to build organizational innovation capacity in a knowledge broker/knowledge curator capacity.

[…] journal club post on this topic.  Furthermore, knowledge mobilization is not challenged by attribution which is an issue in knowledge translation, translation and […]

[…] time.  Annete Boaz says co-production of knowledge can produce an impact on research collaborators even before research is finished.  However, knowledge impact is often a process of gradual enlightenment that can take months or […]

Leave a Comment