Beyond the artifact: rethinking the research output

Many interventions to increase research impact have received a lot of attention recently, including:

• Liberating research from PDFs;
• Changing the writing style from one that is targeted solely at researchers to other audiences;
• Using a variety of techniques to make a research output more engaging, using tools such as Atavist.

But I would argue that we still look at presenting research in far too static a manner, and in fact where we are more concerned about admiring our own handiwork than truly thinking about research more dynamically.+

We need to be more dynamic in three ways:

• In usage, allowing our audiences to use the research in multiple ways. We still assume that our audiences will simply read or skim the report as it is presented. Even if it presented in a more engaging way, we still expect a quite linear and self-contained approach.
• Over time, taking the long view that the specific research output fits within a much longer time horizon of research threads. The current approach is quite static, optimizing the research output for the contemporary trends rather than as research that fits within a bigger timescale.
• As part of a larger whole, allowing our audiences to use a particular research output as a launch point rather than the end destination. Too often research is written in a way that assumes that everyone will want this current gem we are working on when that may not be the case (although people may first arrive on a particular piece of research).

In short, we are a bit too narcissistic about research outputs, polishing them to look good right now without much consideration of the longer view and assuming that people will like it the way we present it (since we like the way it looks). In other words, we are creating artifacts for historical observation later rather than a mesh of research that is toward allowing our audiences to take action.

Allow more dynamic usage

Maybe the goal isn’t to get people to read our research at all—at least the full report, in whatever form it might be.

Obviously we need other researchers to read the full research reports, but most audiences care more about the implications of the research.

Think tanks need to provide actionable research in two ways: 1) pull out more actionable implications and 2) create components that can be used separately.

Although research teams may resist, highlighting actionable implications is important to most audiences. For instance, if a primary audience is policymakers then perhaps key speaking points (at least on key content types) is the most important — of course the full information is there if someone wants to dig in but the most actionable information appears right off the bat. Other actionable outcomes may be things like key findings and recommendations.

Ideally these implications of the research are presented in a way that the entire report does not have to be read, regardless of format of the full report. In some ways, this may mean that a fairly boring presentation (for example with the key findings highlighted on a web page and the entire report is only available in PDF) may be more effective than a report presentation that’s more dynamic.

Key audiences may only want to use parts of your research, for instance data, tables, charts, quotes, or images. Ideally these are developed in a way that allows them to be shared, with a clear way of people to get back to the original research even if the element of the research gets separated from the main research text. For instance, see the Heritage Foundation example (taken from their Issue Brief 4594) below — if this table is separated from the article then someone could still find the full research report (in this example note the “IB 4594 heritage.org” at the bottom of the table — the table user could find the article by searching on “IB 4594” on the Heritage site).

Take the long view

We spend too much time optimizing each particular research output instead of getting the structures in place to support all our research for the long term.

Fashion is fickle. The deeper we customize a particular article toward what impresses a contemporary audience means the more ridiculous it will look later (just like the more fashionable we are the more ridiculous we look later).

But looks aren’t the real problem. The problem is that the more that we customize for the now means the less likely the research will be useful for long. Research is long term. The particular research output isn’t the first and won’t be the last on the research thread.

This occurs in three ways:

1. Over-customization which means that presence-wide changes cannot be rolled out. One problem with over-customization is that if you discover better ways of presenting information they cannot be rolled out to all the existing research. For instance, if you discover that there’s a better way to present data tables but everyone has implemented them in their own way in reports then it can’t be rolled out site-wide.
2. Not using templates that allow pertinent newer research to be highlighted. Sometimes your research will be discovered far in the future. There should be ways of linking into newer research.
3. Overlooking opportunities to weave a longer-term story. Ideally there are ways of presenting this research within the longer term context, in a dynamic way.

View your research as a launch point

The particular research output should not be considered solely as a destination but as a launch point to more about your research.

This may seem counter-intuitive, but your site visitor may arrive on your beautiful report and it may not be what they are most interested in. After all, the report may happen to be the first content of yours that they have seen, and may arrive somewhat arbitrarily, for instance from a Google search.

Ideally you create your research article to include a rich context that allows the ability to go broader (for instance, to see all the research you have on the topic), laterally (to related research), or deeper (to the raw data or the details on the research methodology). This should all be done dynamically, such that as more research is added the contextual information updates as well (this is particularly relevant for the lateral navigation options).