Author: John Borghi

The Significance of Managing Research Data

Some of the most influential research tools of the last century were created to ensure the quality of beer and extrapolate the results of agriculture experiments conducted in the English countryside. Though ostensibly about the placement of a decimal point, an ongoing debate about the application of these tools also provides a window for understanding what it actually means to manage research data.

The p-value: A very quick introduction

Though now ubiquitous in experiment-based research, statistical techniques for extending inferences from small sample (e.g. the participants in a research study) to larger populations are actually a relatively recent invention. The t-test, an early and still widely used example of “small sample” statistics was developed by William Sealy Gossett in the early 20th century as an economical way of ensuring the quality of stout. Several years later, while assisting with long-term experiments on wheat and grass at Rothamsted Experimental Station, Ronald Fisher would build on the work of Gosset and others to develop a statistical framework based around the idea of comparing observations to the null hypothesis- the position that there is no significant difference between two or more specified sets of observations.

In Fisher’s significance testing framework, devices like t-tests are tests of the null hypothesis. The results of these tests indicate the likelihood of observing a result when the null hypothesis is true. The logic is a little tricky, but the core idea is that these tests give researchers a way of understanding the likelihood that their data is the result of sampling or experimental error. In quantitative terms, this likelihood is known as a p-value. In his highly influential 1925 book, Statistical Methods for Research Workers, Fisher would introduce an informal threshold for rejecting the null hypothesis: p < 0.05.

In one of the most influential sentences in modern research methodology, Ronald Fisher describes p = 0.05 as a convenient point for judging the significance of a statistical test. From: Fisher, R.A. (1925). Statistical Methods for Research Workers.

Despite the vehement objections of all three, Fisher’s work would later be synthesized with that of statisticians Jerzy Neyman and Egon Pearson into a suite of tools that are still widely used in many fields of research. In practice, p < 0.05 has since become a one-size-fits-all indicator of success. For decades it has been acknowledged that work that meets this criterion is generally more likely to be reported in the scholarly literature while work that doesn’t is generally relegated the proverbial file drawer.

Beyond p < 0.05

The p < 0.05 threshold has become a flashpoint the ongoing conversation about research practices, reproducibility, and replicability. Heated conversations about the use and misuse of p-values have been ongoing for decades, but over the summer a group of 72 influential researchers proposed a seemingly simple step forward- change the threshold from 0.05 to 0.005. According to the authors, “Reducing the p-value threshold for claims of new discoveries to 0.005 is an actionable step that will immediately improve reproducibility.”.

As of this writing, two responses have been published. Both weigh the pros and cons of p < 0.005 and argue that the placement of a decimal point is less of a problem than the uncritical use of a single one-size-fits-all threshold across many different circumstances and fields of research. Both end on calls for greater transparency and stronger justifications for how decisions related to research design and statistical practice are made. If the initial paper proposed changing the answer from p < 0.05 to 0.005, both responses highlight the necessity of changing the question from one that is focused on statistics to one that incorporates research data management (RDM).

Ensuring that data can be used and evaluated in the future is one of the primary goals of RDM. For example, the RDM guide we’re developing does not have a space for assessing p-values. Instead, its focus is assessing and advancing practices related to planning for, saving, and documenting data and other research products. Such practices come with their own nuance, learning curves, and jargon, but are important elements to any effort to ensure that research decisions are transparent and justified.

Resources and Additional Reading

Benjamin, D. J., Berger, J. O., Johannesson, M., Nosek, B. A., Wagenmakers, E. J., Berk, R., … & Cesarini, D. (2017). Redefine statistical significance. Nature Human Behaviour. doi: 10.1038/s41562-017-0189-z

Lakens, D., Adolfi, F. G., Albers, C. J., Anvari, F., Apps, M. A. J., Argamon, S. E., … Zwaan, R. A. (2017). Justify your alpha: A response to “Redefine statistical significance”PsyArxiv preprint. doi: 10.17605/OSF.IO/9S3Y6

McShane, B. B., Gal, D., Gelman, A., Robert, C., & Tackett, J. L. (2017). Abandon statistical significance. arXiv preprint. arXiv: 1709.07588.

Sterling, T. D. (1959). Publication decisions and their possible effects on inferences drawn from tests of significance—or vice versaJournal of the American Statistical Association54(285), 30-34. doi: 10.1080/01621459.1959.10501497

Rosenthal, R. (1979). The file drawer problem and tolerance for null resultsPsychological Bulletin86(3), 638-641. doi: 10.1037/0033-2909.86.3.638

Managing the new NIH requirements for clinical trials

As part of an effort to enhance transparency in biomedical research, the National Institutes of Health (NIH) have, over the last few years, announced a series of policy changes related to clinical trials. Though there is still a great deal of uncertainty about which studies do and do not qualify, these changes may have significant consequences for researchers who may not necessarily consider their work to be clinical or part of a trial.

Last September, the NIH announced a series of requirements for studies that meet the agency’s revised and expanded definition of a clinical trials. Soon after, it was revealed that many of these requirements may apply to large swaths of NIH-funded behavioral, social science, and neuroscience research that, historically, have not been considered to be clinical in nature. This was affirmed several weeks ago when the agency released a list of case studies that included a brain imaging study in which healthy participants completed a memory task as an example of a clinical trial.


NIH’s revised and expanded definition of clinical trials includes many approaches to human subjects research that have historically been considered basic research. (Source)

What exactly constitutes a clinical trial now?

Because many investigators doing behavioral, social science, and neuroscience research consider their work to be basic research and not a part of a clinical trial, it is worth taking a step back to consider how NIH now defines the term.

According to the NIH, clinical trials are “studies involving human participants assigned to an intervention in which the study is designed to evaluate the effect(s) of the intervention on the participant and the effect being evaluated is a health-related biomedical or behavioral outcome.”, In an NIH context, intervention refers to “a manipulation of the subject or subject’s environment for the purpose of modifying one or more health-related biomedical or behavioral processes and/or endpoints.”. Because the agency considers all of the studies it funds that investigate biomedical or behavioral outcomes to be health-related, this definition includes mechanistic or exploratory work that does not have direct clinical implications.

Basically, if you are working on an NIH-funded study that involves biomedical or behavioral variables, you should be paying attention to the new requirements about clinical trials.

What do I need to do now that my study is considered a clinical trial?

If you think your work may be reclassified as a clinical trial, it’s probably worth getting a head start on meeting the new requirements. Here is some practical advice about getting started.


The new NIH requirements for clinical trials affect activity throughout the lifecycle of a research project. (Source)

Applying for Funding

NIH has specified new requirements about how research involving clinical trials can be funded. For example, NIH will soon require that any application involving a clinical trial be submitted in response to a funding opportunity announcement (FOA) or request for proposal (RFP) that explicitly states that it will accept a clinical trial. This means, that if you are a researcher whose work involves biomedical or behavioral measures, you may have to apply to funding mechanisms that your peers have argued are not necessarily optimal or appropriate. Get in touch with your program officer and watch this space.

Grant applications will also feature a new form that consolidates the human subjects and clinical trial information previously collected across multiple forms into one structured form. For a walkthrough of the new form, check out this video.

Human Subjects Training

Investigators involved in a clinical trial must complete Good Clinical Practice (GCP) training. GCP training addresses elements related to the design, conduct, and reporting of clinical trials and can be completed via a class or course, academic training program, or certification from a recognized clinical research professional organization.

In practice, if you have already completed human subjects training (e.g. via CITI) and believe your research may soon be classified as a clinical trials, you may want to get proactive about completing those couple additional modules.

Getting IRB Approval

Good news if you work on a multi-site study, NIH now expects that you will use a single Institutional Review Board (sIRB) for ethical review. This should help streamline the review process, since it will no longer be necessary to submit an application to each site’s individual IRB. This requirement also applies to studies that are not clinical trials.

Registration and Reporting

NIH-funded projects involving clinical trials must be registered on In practice, this means that the primary investigator or grant awardee is responsible for registering the trial no later than 21 days after the enrollment of the first participant and is required to submit results information no later than a year after the study’s completion date. Registration involves supplying a significant amount of information about a study’s planned design and participants while results reporting involves supplying information about the participants recruited, the data collected, and the statistical tests applied. For more information about, check out this paper.

If you believe your research may soon be reclassified as a clinical trial, now is probably a good time to take a hard look at how you and your lab handle research data management.The best way to relieve the administrative burden of these new requirements is to plan ahead and ensure that your materials are well organized, your data is securely saved, and your decisions are well documented. The more you think through how you’re going to manage your data and analyses now, the less you’ll have to scramble to get everything together when the report is due. If you haven’t already, now would be a good time to get in touch with the data management, scholarly communications, and research IT professionals at your institution.

What We Talk About When We Talk About Reproducibility

At the very beginning of my career in research I conducted a study which involved asking college students to smile, frown, and then answer a series of questions about their emotional experience. This procedure was based on several classic studies which posited that, while feeling happy and sad makes people smile and frown, smiling and frowning also makes people feel happy and sad. After several frustrating months of trying and failing to get this to work, I ended my experiment with no significant results. At the time, I chalked up my lack of success to inexperience. But then, almost a decade later, a registered replication report of the original work also showed a lack of significant results and I was left to wonder if I had also been caught up in what’s come to be known as psychology’s reproducibility crisis.


Campbell’s Soup Cans (1962) by Andy Warhol. Created by replicating an existing object and then reproducing the process at least 32 times.

While I’ve since left the lab for the library, my work still often intersects with reproducibility. Earlier this year I attended a Research Transparency and Reproducibility Training session offered by the Berkeley Institute for Transparency in the Social Sciences (BITSS) and my projects involving brain imaging data, software, and research data management all invoke the term in some way.  Unfortunately, though it has always has been an important part of my professional activities, it isn’t always clear to me what we’re actually talking about when we talk about reproducibility.

The term “reproducibility” has been applied to efforts to enhance or ensure the research process for at at least 25 years. However, related conversations about how research is conducted, published, and interpreted have been ongoing for more than half a century. Ronald Fisher, who popularized the p-value that lies so central to many modern reproducibility efforts, summed up the situation in 1935.

“We may say that a phenomenon is experimentally demonstrable when we know how to conduct an experiment which will rarely fail to give us statistically significant results.”

Putting this seemingly simple statement into action has proven to be quite complex. Some reproducibility-related efforts are aimed at how researchers share their results, others are aimed at how they define statistical significance. There is now a burgeoning body of scholarship devoted to the topic. Even putting aside terms like HARKing, QRPs, and p-hacking, seemingly mundane objects like file drawers are imbued with particular meaning in the language of reproducibility.

So what actually is reproducibility?

Well… it’s complicated.

The best place to start might be the National Science Foundation, which defines reproducibility as “The ability of a researcher to duplicate the results of a prior study using the same materials and procedures used by the original investigator.”. According the NSF, reproducibility is one of three qualities that ensure research is robust. The other two, replicability and generalizability, are defined as “The ability of a researcher to duplicate the results of a prior study if the same procedures are followed but new data are collected.” and “Whether the results of a study apply in other contexts or populations that differ from the original one.” respectively. The difference between these terms is in the degree of separation from the original research, but all three converge on the quality of research. Good research is reproducible, replicable, and generalizable and , at least in the context of the NSF, a researcher invested in ensuring the reproducibility of their work would deposit their research materials and data in a manner and location where they could be accessed and used by others.

Unfortunately, defining reproducibility isn’t always so simple. For example, according to the NSF’s terminology, the various iterations of the Reproducibility Project are actually replicability projects (muddying the waters further, the Reproducibility Project: Psychology was preceded by the Many Labs Replication Project). However, the complexity of defining reproducibility is perhaps best illustrated by comparing the NSF definition to that of the National Institutes of Health.

Like the NSF, NIH invokes reproducibility in the context of addressing the quality of research. However, unlike the NSF, the NIH does not provide an explicit definition of the term. Instead NIH grant applicants are asked to address rigor and reproducibility across four areas of focus: scientific premise, scientific rigor (design), biological variables, and authentication. Unlike the definition supplied by the NSF, NIH’s conception of reproducibility appears to apply to an extremely broad set of circumstances and encompasses both replicability and generalizability. In the context of the NIH, a researcher invested in reproducibility must critically evaluate every aspect of their research program to ensure that any conclusions drawn from it are well supported.

Beyond the NSF and NIH, there have been numerous attempts to clarify what reproducibility actually means. For example, a paper out of the Meta-Research Innovation Center at Stanford (METRICS) distinguishes between “methods reproducibility”, “results reproducibility”, and “inferential reproducibility”. Methods and results reproducibility map onto the NSF definitions of reproducibility and replicability, while inferential reproducibility includes the NSF definition of generalizability and also the notion of different researchers reaching the same conclusion following reanalysis of the original study materials. Other approaches focus on methods by distinguishing between empirical, statistical, and computational reproducibility or specifying that replications can be direct or conceptual.

No really, what actually is reproducibility?

It’s everything.

The deeper we dive into defining “reproducibility”, the muddier the waters become. In some contexts, the term refers to very specific practices related to authenticating the results of a single experiment. In other contexts, it describes a range of interrelated issues related to how research is conducted, published, and interpreted. For this reason, I’ve started to move away from explicitly invoking the term when I talk to researchers. Instead, I’ve tried to frame my various research and outreach projects in terms of how they relate to fostering good research practice.

To me, “reproducibility” is about problems. Some of these problems are technical or methodological and will evolve with the development of new techniques and methods. Some of these problems are more systemic and necessitate taking a critical look at how research is disseminated, evaluated, and incentivized. But fostering good research practice is central to addressing all of these problems.

Especially in my current role, I am not particularly well equipped to speak to if a researcher should define statistical significance as p < 0.05, p < 0.005, or K > 3. What I am equipped to do is to help a researcher manage their research materials so they can be used, shared, and evaluated over time. It’s not that I think the term is not useful, but the problems conjured by reproducibility are so complex and context dependent that I’d rather just talk about solutions.

Resources for understanding reproducibility and improving research practice

Goodman A., Pepe A, Blocker A. W., Borgman C. L., Cranmer K., et al. (2014) Ten simple rules for the care and feeding of scientific data. PLOS Computational Biology 10(4): e1003542.

Ioannidis J. P. A. (2005) Why most published research findings are false. PLOS Medicine 2(8): e124.

Kitzes, J., Turek, D., & Deniz, F. (Eds.). (2017). The Practice of Reproducible Research: Case Studies and Lessons from the Data-Intensive Sciences. Oakland, CA: University of California Press.

Munafò, M. R., Nosek, B. A., Bishop, D. V., Button, K. S., Chambers, C. D., du Sert, N. P., et al. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1, 0021.

Wilson Gl, Bryan J., Cranston K., Kitzes J., Nederbragt L., et al. (2017) Good enough practices in scientific computing. PLOS Computational Biology 13(6): e1005510.

From Brain Blobs to Research Data Management

If you spend some time browsing the science section of a publication like the New York Times you’ll likely run across an image that looks something like the one below: A cross section of a brain covered in colored blobs. These images are often used to visualize the results of studies using a technique called functional magnetic resonance imaging (fMRI), a non-invasive method for measuring brain activity (or, more accurately, a correlate of brain activity) over time. Researchers who use fMRI are often interested in measuring the activity associated with a particular mental process or clinical condition.


A visualization of the results of an fMRI study. These images are neat to look at but not particularly useful without information the underlying data and analysis.

Because of the size and complexity of the datasets involved, research data management (RDM) is incredibly important in fMRI research. In addition to the brain images, a typical fMRI study involves the collection of questionnaire data, behavioral measures, and sensitive medical information. Analyzing all this data often requires the development of custom code or scripts. This analysis is also iterative and cumulative, meaning that a researcher’s decisions at each step along the way can have significant effects on both the subsequent steps and what is ultimately reported in a presentation, poster, or journal article. Those blobby brain images may look cool, but they aren’t particularly useful in the absence of information about the underlying data and analyses.

In terms of both the financial investment and researcher hours involved, fMRI research is quite expensive. Throughout fMRI’s relatively short history, data sharing has been proposed multiple times times as a method for maximizing the value of individual datasets and for overcoming the field’s ongoing methodological issues. Unfortunately, a very practical issue has hampered efforts to foster the open sharing of fMRI data- researchers have historically organized, documented, and saved their data (and code) in very different ways.

What we are doing and why

Recently, following concerns about sub-optimal statistical practices and long-standing software errors, fMRI researchers have begun to cohere around a set of standards regarding how data should be collected, analyzed, and reported. From a research data management perspective, it’s also very exciting to see that there is also an emerging standard regarding how data should be organized and described. But, even with these emerging standards, our understanding of the data-related practices actually employed by fMRI in the lab and how those practices relate to data sharing and other open science-related activities remains mostly anecdotal.

To help fill this knowledge gap and hopefully advance some best practices related to data management and sharing, Dr. Ana Van Gulick and I are conducting a survey of fMRI researchers. Developed in consultation with members of the open and reproducible neuroscience communities, our survey asks researchers about their own data-related practices, how they view the field as a whole, their interactions with RDM service providers, and the degree to which they’ve embraced developments like registrations and pre-prints. Our hope is that our results will be useful for both the community of researchers who use fMRI but and for data service providers looking to engage with researchers on their own terms.

If you are a researcher who uses fMRI and would like to complete our survey, please follow this link. We estimate that the survey should take between 10 and 20 minutes.

If you are a data service provider and would like to chat with us about what we’re doing and why, please feel free to either leave a comment or contact me directly.

Building a Community: Three months of Library Carpentry.

Back in May, almost 30 librarians, researchers, and faculty members got together in Portland Oregon to learn how to teach lessons from Software, Data, and Library Carpentry. After spending two days learning the ins and outs of Carpentry pedagogy and live coding, we all returned to our home institutions, as part of the burgeoning Library Carpentry community.

Library Carpentry didn’t begin in Portland, of course. It began in 2014 when the community began developing a group of lessons at the British Library. Since then, dozens of Library Carpentry workshops have been held across four continents. But the Portland event, hosted by California Digital Library, was the first Library Carpentry-themed instructor training session. Attendees not only joined the Library Carpentry community, but took their first step in getting certified as Software and Data Carpentry instructors. If Library Carpentry was born in London, it went through a massive growth spurt in Portland.

Together, the carpentries are a global movement focused on teaching people computing skills like navigating the Unix Shell, doing version control with Git, and programming with Python. While Software and Data Carpentry are focused on researchers, Library Carpentry is by and for Librarians. Library Carpentry lessons include an introduction to data for librarians, Open Refine, and many more. Many attendees of the Portland instructor training contributed to these lessons during the Mozilla Global Sprint in June. After more than 850 Github events (pull requests, forks, issues, etc), Library Carpentry ended up as far and away the most active part of the global sprint. We even had a five month old get in on the act!

Since the instructor training and the subsequent sprint, a number of Portland attendees have completed their instructor certification. We are on track to have 10 certified instructors in the UC system alone. Congratulations, everyone!

Building an RDM Guide for Researchers – An (Overdue) Update

It has been a little while since I last wrote about the work we’re doing to develop a research data management (RDM) guide for researchers. Since then, we’ve thought a lot about the goals of this project and settled on a concrete plan for building out our materials. Because we will soon be proactively seeking feedback on the different elements of this project, I wanted to provide an update on what we’re doing and why.


A section of the Rosetta Stone. Though it won’t help decipher Egyptian hieroglyphs, we hope our RDM guide will researchers and data service providers speak the same language. Image from the British Museum.

Communication Barriers and Research Data Management

Several weeks ago I wrote about addressing Research Data Management (RDM) as a “wicked problem”, a problem that is difficult to solve because different stakeholders define and address it in different ways. My own experience as a researcher and library postdoc bears this out. Researchers and librarians often think and talk about data in very different ways! But as researchers face changing expectations from funding agencies, academic publishers, their own peers, and other RDM stakeholders about how they should manage and share their data, overcoming such communication barriers becomes increasingly important.

From visualizations like the ubiquitous research data lifecycle to instruments like the Data Curation Profiles, there are a wide variety of excellent tools that can be used to facilitate communication between different RDM stakeholders. Likewise, there are also discipline-specific best practice guidelines and tools like the Research Infrastructure Self Evaluation Framework (RISE) that allow researchers and organizations to assess and advance their RDM activities. What’s missing is a tool that combines these two elements that enables researchers the means to easily self-assess where they are in regards to RDM and allows data service providers to provide easily customizable guidance about how to advance their data-related practices.

Enter our RDM guide for researchers.

Our RDM Guide for Researchers

What I want to emphasize most about our RDM guide is that it is, first and foremost, designed to be a communication tool. The research and library communities both have a tremendous amount of knowledge and expertise related to data management. Our guide is not intended to supplant tools developed by either, but to assist in overcoming communication barriers in a way that removes confusion, grows confidence, and helps people in both communities find direction.

While the shape of RDM guide has not changed significantly since my last post, we have refined its basic structure and have begun filling in the details.

The latest iteration of our guide consists of two main elements:

  1. A RDM rubric which allows researchers to self-assess their data-related practices using language and terminology with which they are familiar.
  2. A series of one page guides that provide information about how to advance data-related practices as necessary, appropriate, or desired.
RDM_rubric (1)

The two components of our RDM Guide for Researchers. The rubric is intended to help researchers orient themselves in the ever changing landscape of RDM while the guides are intended to help them move forward.

The rubric is similar to the “maturity model”  described in my earlier blog posts. In this iteration, it consists of a grid containing three columns and a number of rows. The leftmost column contains descriptions of different phases of the research process. At present, the rubric contains four such phases: Planning, Collection, Analysis, and Sharing. These research data lifecycle-esque terms are in place to provide a framing familiar to data service providers in the library and elsewhere.

The next column includes phrases that describe specific research activities using language and terminology familiar to researchers. The language in this column is, in part, derived from the unofficial survey we conducted to understand how researchers describe the research process. By placing these activities beside those drawn from the research data lifecycle, we hope to ground our model in terms that both researchers and RDM service providers can relate to.

The rightmost column then contains a series of declarative statements which a researcher can use to identify their individual practices in terms of the degree to which they are defined, communicated, and forward thinking.

Each element of the rubric is designed to be customizable. We understand that RDM service providers at different institutions may wish to emphasize different services toggled to different parts data lifecycle and that researchers in different disciplines may have different ways of describing their data-related activities. For example, while we are working on refining the language of the declarative statements, I have left them out of the diagram above because they are likely the  rubric that will remain most open for customization.

Each row within the rubric will be complemented by a one page guide that will provide researchers with concrete information about data-related best practices. If the purpose of the rubric is to allow researchers to orient themselves in RDM landscape, the purpose of these guides is to help them move forward.

Generating Outputs

Now that we’ve refined the basic structure of our model, it’s time to start creating some outputs. Throughout the remainder of the summer and into the autumn, members of the UC3 team will be meeting regularly to review the content of the first set of one page guides. This process will inform our continual refinement of the RDM rubric which will, in turn, shape the writing of a formal paper.

Moving forward, I hope to workshop this project with as many interested parties as I can, both to receive feedback on what we’ve done so far and to potentially crowdsource some of the content. Over the next few weeks I’ll be soliciting feedback on various aspects of the RDM rubric. If you’d like to provide feedback, please either click through the links below (more to be added in the coming weeks) or contact me directly.


Provide feedback on our guide!

Planning for Data

More coming soon!

Talking About Data: Lessons from Science Communication

As a person who worked for years in psychology and neuroscience laboratories before coming to work in academic libraries, I have particularly strong feelings about ambiguous definitions. One of my favorite anecdotes about my first year of graduate school involves watching two researchers argue about the definition of “attention” for several hours, multiple times a week, for an entire semester. One of the researchers was a clinical psychologist, the other a cognitive psychologist. Though they both devised research projects and wrote papers on the topic of attention, their theories and methods could not have been more different. The communication gap between them was so wide that they were never able to move forward productively. The punchline is that, after sitting through hours of their increasingly abstract and contentious arguments, I would go on to study attention using yet another set of theories and methods as a cognitive neuroscientist. Funny story aside, this anecdote illustrates the degree to which people with different perspectives and levels of expertise can define the same problem in strikingly different ways.


A facsimile of a visual search array used by cognitive psychologists to study attention. Spot the horizontal red rectangle.

In the decade that has elapsed since those arguments, I have undergone my own change in perspective- from a person who primarily collects and analyzes their own research data to a person who primarily thinks about ways to help other researchers manage and share their data. While my day-to-day activities look rather different, there is one aspect of my work as a library post-doc that is similar to my work as a neuroscientist- many of my colleagues ostensibly working on the same things often have strikingly different definitions, methods, and areas of expertise. Fortunately, I have been able to draw on a body of work that addresses this very thing- science communication.

Wicked Problems

A “wicked problem” is a problem that is extremely difficult to solve because different stakeholders define and address it in different ways. In my anecdote about argumentative professors, understanding attention can be considered a wicked problem. Without getting too much into the weeds, the clinical psychologist understood attention mostly in the context of diagnoses like Attention Deficit Disorder, while the cognitive psychologist understood it the context of scanning visual environments for particular elements or features. As a cognitive neuroscientist, I came to understand it mostly in terms of its effects within neural networks as measured by brain imaging methods like fMRI.

Research data management (RDM) has been described as a wicked problem. A data service provider in an academic library may define RDM as “the documentation, curation, and preservation of research data”, while a researcher may define RDM as either simply part of their daily work or, in the case of something like a data management plan written for a grant proposal, as an extra burden placed upon such work. Other RDM stakeholders, including those affiliated with IT, research support, and university administration, may define it in yet other ways.

Science communication is chock full of wicked problems, including concepts like climate change and the use of stem cell use. Actually, given the significant amount of scholarship devoted to defining terms like “scientific literacy” and the multitudes of things that the term describes, science communication may itself be a wicked problem.

What is Scientific Communication?

Like attention and RDM, it is difficult to give a comprehensive definition of science communication. Documentaries like “Cosmos” are probably the most visible examples, but science communication actually comes in a wide variety of forms including science journalism, initiatives aimed at science outreach and advocacy, and science art. What these activities have in common is that they all generally aim to help people make informed decisions in a world dominated by science and technology. In parallel, there is also a burgeoning body of scholarship devoted to the science of science communication which, among other things, examines how effective different communication strategies are for changing people’s perceptions and behaviors around scientific topics.

For decades, the prevailing theory in science communication was the “Deficit Model”, which posits that scientific illiteracy is due to a simple lack of information. In the deficit model, skepticism about topics such as climate change are assumed to be due to a lack of comprehension of the science behind them. Thus, at least according to the deficit model, the “solution” to the problem of science communication is as straightforward as providing people with all the facts. In this conception, the audience is generally assumed to be homogeneous and communication is assumed to be one way (from scientists to the general public).

Though the deficit model persists, study after study (after meta-analysis) has shown that merely providing people with facts about a scientific topic does not cause them to change their perceptions or behaviors related to that topic. Instead, it turns out that presenting facts that conflict with a person’s worldview can actually cause them to double down on that worldview. Also, audiences are not homogenous. Putting aside differences in political and social worldviews, people have very different levels of scientific knowledge and relate to that knowledge in very different ways. For this reason, more modern models of science communication focus not on one-way transmissions of information but on fostering active engagement, re-framing debates, and meeting people where they are. For example, one of the more effective strategies for getting people to pay attention to climate change is not to present them with a litany of (dramatic and terrifying) facts, but to link it to their everyday emotions and concerns.


Find the same rectangle as before. It takes a little longer now that the other objects have a wider variety of features, right? Read more about visual search tasks here.

Communicating About Data

If we adapt John Durant’s nicely succinct definition of science literacy,What the general public ought to know about science.” to an RDM context, the result is something like “What researcher’s out to know about handling data.” Thus, data services in academic libraries can be said to be a form of science communication. As with “traditional” science communicators, data service providers interact with audiences possessing different perspectives and levels of knowledge as their own. The major difference, of course, being that the audience for data service providers is specifically the research community.

There is converging evidence that many of the current plans for fostering better RDM have led to mixed results. Recent studies of NSF data management plans have revealed a significant amount of variability in terms of the degree to which researchers address data management-related concepts like metadata, data sharing, and long-term preservation. The audience of data service providers is, like those of more “traditional science communicators, quite heterogeneous, so perhaps adopting methods from the repertoire of science communication could help foster more active engagement and the adoption of better practices. Many libraries and data service providers have already adopted some of these methods, perhaps without realizing their application in other domains. But I also don’t mean to criticize any existing efforts to engage researchers on the topic of RDM. If I’ve learned one thing from doing different forms of science communication over the years, it is that outreach is difficult and change is slow.

In a series of upcoming blog posts, I’ll write about some of my current projects that incorporate what I’ve written here. First up: I’ll provide an update of the RDM Maturity Model project that I previously described here and here. Coming soon!

Understanding researcher needs and values related to software

Software is as important as data when it comes to building upon existing scholarship. However, while there has been a small amount of research into how researchers find, adopt, and credit it, there is a comparative lack of empirical data on how researchers use, share, and value their software.

The UC Berkeley Library and the California Digital Library are investigating researchers’ perceptions, values, and behaviors in regards to software generated as part of the research process. If you are a researcher, it would be greatly appreciated if you could spare 10-15 minutes to complete the following survey:

Take the survey now!

The results of this survey will help us better understand researcher needs and values related to software and may also inform the development of library services related to software best practices, code sharing, and the reproducibility of scholarly activity.

If you have questions about our study or any problems accessing the survey, please contact or

An RDM Model for Researchers: What we’ve learned

Thanks to everyone who gave feedback on our previous blog post describing our data management tool for researchers. We received a great deal of input related to our guide’s use of the term “data sharing” and our guide’s position in relation to other RDM tools as well as quite a few questions about what our guide will include as we develop it further.

As stated in our initial post, we’re building a tool to enable individual researchers to assess the maturity of their data management practices within an institutional or organizational context. To do this, we’ve taken the concept of RDM maturity from in existing tools like the Five Organizational Stages of Digital Preservation, the Scientific Data Management Capability Model, and the Capability Maturity Guide and placed it within a framework familiar to researchers, the research data lifecycle.


A visualization of our guide as presented in our last blog post. An updated version, including changed made in response to reader feedback, is presented later in this post.

Data Sharing

The most immediate feedback we received was about the term “Data Sharing”. Several commenters pointed out the ambiguity of this term in the context of the research data life cycle. In the last iteration of our guide, we intended “Data Sharing” as a shorthand to describe activities related to the communication of data. Such activities may range from describing data in a traditional scholarly publication to depositing a dataset in a public repository or publishing a data paper. Because existing data sharing policies (e.g. PLOS, The Gates Foundation, and The Moore Foundation) refer specifically to the latter over the former, the term is clearly too imprecise for our guide.

Like “Data Sharing”, “Data Publication” is a popular term for describing activities surrounding the communication of data. Even more than “Sharing”, “Publication” relays our desire to advance practices that treat data as a first class research product. Unfortunately the term is simultaneously too precise and too ambiguous it to be useful in our guide. On one hand, the term “Data Publication” can refer specifically to a peer reviewed document that presents a dataset without offering any analysis or conclusion. While data papers may be a straightforward way of inserting datasets into the existing scholarly communication ecosystem, they represent a single point on the continuum of data management maturity. On the other hand, there is currently no clear consensus between researchers about what it means to “publish” data.

For now, we’ve given that portion of our guide the preliminary label of “Data Output”. As the development process proceeds, this row will include a full range of activities- from description of data in traditional scholarly publications (that may or may not include a data availability statement) to depositing data into public repositories and the publication of data papers.

Other Models and Guides

While we correctly identified that there are are range of rubrics, tools, and capability models with similar aims as our guide, we overstated that ours uniquely allows researchers to assess where they are and where they want to be in regards to data management. Several of the tools we cited in our initial post can be applied by researchers to measure the maturity of data management practices within a project or institutional context.

Below we’ve profiled four such tools and indicated how we believe our guide differs from each. In differentiating our guide, we do not mean to position it strictly as an alternative. Rather, we believe that our guide could be used in concert with these other tools.

Collaborative Assessment of Research Data Infrastructure and Objectives (CARDIO)

CARDIO is a benchmarking tool designed to be used by researchers, service providers, and coordinators for collaborative data management strategy development. Designed to be applied at a variety of levels, from entire institutions down to individual research projects, CARDIO enables its users to collaboratively assess data management requirements, activities, and capacities using an online interface. Users of CARDIO rate their data management infrastructure relative to a series of statements concerning their organization, technology, and resources. After completing CARDIO, users are given a comprehensive set of quantitative capability ratings as well as a series of practical recommendations for improvement.

Unlike CARDIO, our guide does not necessarily assume its users are in contact with data-related service providers at their institution. As we stated in our initial blog post, we intend to guide researchers to specialist knowledge without necessarily turning them into specialists. Therefore, we would consider a researcher making contact with their local data management, research IT, or library service providers for the first time as a positive application of our guide.

Community Capability Model Framework (CCMF)

The Community Capability Model Framework is designed to evaluate a community’s readiness to perform data intensive research. Intended to be used by researchers, institutions, and funders to assess current capabilities, identify areas requiring investment, and develop roadmaps for achieving a target state of readiness, the CCMF encompasses eight “capability factors” including openness, skills and training, research culture, and technical infrastructure. When used alongside the Capability Profile Template, the CCMF provides its users with a scorecard containing multiple quantitative scores related to each capability factor.

Unlike the CCMF, our guide does not necessarily assume that its users should all be striving towards the same level of data management maturity. We recognize that data management practices may vary significantly between institutions or research areas and that what works for one researcher may not necessarily work for another. Therefore, we would consider researchers understanding the maturity of their data management practices within their local contexts to be a positive application of our guide.

Data Curation Profiles (DCP) and DMVitals

The Data Curation Profile toolkit is intended to address the needs of an individual researcher or research group with regards to the “primary” data used for a particular project. Taking the form of a structured interview between an information professional and a researcher, a DCP can allow an individual research group to consider their long-term data needs, enable an institution to coordinate their data management services, or facilitate research into broader topics in digital curation and preservation.

DMVitals is a tool designed to take information from a source like a Data Curation Profile and use it to systematically assess a researcher’s data management practices in direct comparison to institutional and domain standards. Using the DMVitals, a consultant matches a list of evaluated data management practices with responses from an interview and ranks the researcher’s current practices by their level of data management “sustainability.” The tool then generates customized and actionable recommendations, which a consultant then provides to the researcher as guidance to improve his or her data management practices.

Unlike DMVitals, our guide does not calculate a quantitative rating to describe the maturity of data management practices. From a measurement perspective, the range of practice maturity may differ between the four stages of our guide (e.g. the “Project Planning” stage could have greater or fewer steps than the “Data Collection” stage), which would significantly complicate the interpretation of any quantitative ratings derived from our guide. We also recognize that data management practices are constantly evolving and likely dependent on disciplinary and institutional context. On the other hand, we also recognize the utility of quantitative ratings for benchmarking. Therefore, if, after assessing the maturity of their data management practices with our guide, a researcher chooses to apply a tool like DMVitals, we would consider that a positive application of our guide.

Our Model (Redux)

Perhaps the biggest takeaway from the response to our  last blog post is that it is very difficult to give detailed feedback on a guide that is mostly whitespace. Below is an updated mock-up, which describes a set of RDM practices along the continuum of data management maturity. At present, we are not aiming to illustrate a full range of data management practices. More simply, this mock-up is intended to show the types of practices that could be described by our guide once it is complete.


An updated visualization of our guide based on reader feedback. At this stage, the example RDM practices are intended to be representative not comprehensive.

Project Planning

The “Project Planning” stage describes practices that occur prior to the start of data collection. Our examples are all centered around data management plans (DMPs), but other considerations at this stage could include training in data literacy, engagement with local RDM services, inclusion of “sharing” in project documentation (e.g. consent forms), and project pre-registration.

Data Collection

The “Data Collection” stage describes practices related to the acquisition, accumulation, measurement, or simulation of data. Our examples relate mostly to standards around file naming and structuring, but other considerations at this stage could include the protection of sensitive or restricted data, validation of data integrity, and specification of linked data.

Data Analysis

The “Data Analysis” stage describes practices that involve the inspection, modeling, cleaning, or transformation of data. Our examples mostly relate to documenting the analysis workflow, but other considerations at this stage could include the generation and annotation of code and the packaging of data within sharable files or formats.

Data Output

The “Data Output” stage describes practices that involve the communication of either the data itself of conclusions drawn from the data. Our examples are mostly related to the communication of data linked to scholarly publications, but other considerations at this stage could include journal and funder mandates around data sharing, the publication of data papers, and the long term preservation of data.

Next Steps

Now that we’ve solicited a round of feedback from the community that works on issues around research support, data management, and digital curation, our next step is to broaden our scope to include researchers.

Specifically we are looking for help with the following:

  • Do you find the divisions within our model useful? We’ve used the research data lifecycle as a framework because we believe it makes our tool user-friendly for researchers. At the same time, we also acknowledge that the lines separating planning, collection, analysis, and output can be quite blurry. We would be grateful to know if researchers or data management service providers find these divisions useful or overly constrained.
  • Should there be more discrete “steps” within our framework? Because we view data management maturity as a continuum, we have shied away from creating discrete steps within each division. We would be grateful to know how researchers or data management service providers view this approach, especially when compared to the more quantitative approach employed by CARDIO, the Capability Profile Template, and DMVitals.
  • What else should we put into our model? Researchers are faced with changing expectations and obligations in regards to data management. We want our model to reflect that. We also want our model to reflect the relationship between research data management and broader issues like openness and reproducibility. With that in mind, what other practices and considerations should or model include?

Building a user-friendly RDM maturity model

UC3 is developing a guide to help researchers assess and progress the maturity of their data management practices.

What are we doing?

Researchers are increasingly faced with new expectations and obligations in regards to data management. To help researchers navigate this changing landscape and to complement existing instruments that enable librarians and other data managers to assess the maturity of data management practices at an institutional or organizational level, we’re developing a guide that will enable researchers to assess the maturity of their individual practices within an institutional or organizational context.

Our aim is to be descriptive rather than prescriptive. We do not assume every researcher will want or need to achieve the same level of maturity for all their data management practices. Rather, we aim to provide researchers with a guide to specialist knowledge without necessarily turning researchers into specialists. We want to help researchers understand where they are and, where appropriate, how to get to where they want or need to be.

Existing Models

As a first step in building our own guide, we’ve researched the range of related tools, rubrics, and capability models. Many, including the Five Organizational Stages of Digital Preservation, the Scientific Data Management Capability Model, and the Capability Maturity Guide developed by the Australian National Data Service, draw heavily from the SEI Capability Maturity Model and are intended to assist librarians, repository managers, and other data management service providers in benchmarking the policies, infrastructure, and services of their organization or institution.  Others, including the Collaborative Assessment of Research Data Infrastructure and Objectives (CARDIO), DMVitals, and the Community Capability Framework, incorporate feedback from researchers and are intended to assist in benchmarking a broad set of data management-related topics for a broad set of stockholders – from organizations and institutions down to individual research groups.

We intend for our guide to build on these tools but to have a different, and we think novel, focus. While we believe it could be a useful tool for data management service providers, the intended audience of our guide is research practitioners. While integration with service providers in the library, research IT, and elsewhere will be included where appropriate, the the focus will be on equipping researchers to assess and refine their individual own data management activities. While technical infrastructure will be included where appropriate, the focus will be on behaviors, “soft skills”, and training.

Our Guide

Below is a preliminary mockup of our guide. Akin to the “How Open Is It?” guide developed by SPARC, PLOS, and the OASPA, our aim is to provide a tool that is comprehensive, user-friendly, and provides tangible recommendations.


Obviously we still have a significant amount of work to do to refine the language and fill in the details. At the moment, we are using elements of the research data lifecycle to broadly describe research activities and very general terms to describe the continuum of practice maturity. Our next step is to fill in the blanks- to more precisely describe research activities and more clearly delineate the stages of practice maturity. From there, we will work to outline the behaviors, skills, and expertise present for each research activity at each stage.

Next Steps

Now that we’ve researched existing tools for assessing data management services and sketched out a preliminary framework for our guide, our next step is to elicit feedback from the broader community that works on issues around research support, data management, and digital curation and preservation.

Specifically we are looking for help on the following:

  • Have we missed anything? There is a range of data management-related rubrics, tools, and capability models – from the community-focused frameworks described above to frameworks focused on the preservation and curation of digital assets (e.g. the Digital Asset Framework, DRAMBORA). As far as we’re aware, there isn’t a complementary tool that allows researchers to assess where they are and where they want to be in regards to data management. Are there efforts that have already met this need? We’d be grateful for any input about the existence of frameworks with similar goals.
  • What would be the most useful divisions and steps within our framework? The “three legged stool” developed by the Digital Preservation Management workshop has been highly influential for community and data management provider-focused tools. Though examining policies, resources, and infrastructure are also important for researchers when self-assessing their data management practices, we believe it would be more useful for our guide to be more reflective of how data is generated, managed, disseminated in a research context. We’d be grateful for any insight into how we could incorporate related models – such as those depicting the research data lifecycle – into our framework.