Data Management Threshold Concepts

We’ve been going through the new ACRL “Framework for Information Literacy for Higher Education” recently at work. This document discusses ways to teach students how to search and understand information resources, framing critical skills as “threshold concepts”. While the Framework itself is interesting, I’m really intrigued by the idea of a threshold concept and wonder if there are any threshold concepts for data management.

For those unfamiliar with the term, a “threshold concept” is an idea that, once understood, completely reframes the way you view a topic. It’s like seeing a hidden image in that it’s very difficult to un-see the image afterward. Threshold concepts are so fundament to understanding that it’s actually necessary to understand the concept in order to progress in the field.

Let’s look at the ACRL Framework to better understand how such concepts work. The six concepts are:

  • Authority is Constructed and Contextual
  • Information Creation as a Process
  • Information Has Value
  • Research as Inquiry
  • Scholarship as Conversation
  • Searching as Strategic Exploration

If you understand these concepts, you’ll easily see, for example, why a scholarly article may be an appropriate source for one research project while a blog post would be better for a different project, depending on the topic. Or why searching doesn’t always turn up the content you are looking for on the first try. Etc.

This blog post is not about the new Framework, but rather how the Framework challenged me to think about what the threshold concepts are for data management. Taking a stab at it (directly cribbing from the Framework), I have three ideas:

  • Data is Contextual
  • Data Management is a Process
  • Data Has Value

Let’s look at these individually to get into what I mean in each case.

First, data is contextual. That means that data never exists independently of the information about how it was acquired and processed. Just like a chemist records notes about her data in a lab notebook, so should any dataset come with enough documentation to be understood by someone who is not the dataset creator. Without this extra information, the data is practically useless.

Second, data management is a process. It’s not something that you do once and are done with forever. It’s a process by which you take better care of your data continually over time. That doesn’t mean that it’s incredibly difficult. Rather, it’s like doing regular preventative maintenance to avoid disaster.

Third, data has value. This is something that many researchers are currently grappling with due to new data sharing requirements. If you can understand that your data has value, you can see how published data adds richness to an article, why data should be preserved after the end of a project, and why other researchers might want to use your data (hint: it’s valuable!).

These three ideas are by no means the final say on threshold concepts in data management, only my initial ideas. I’m still mulling them over (for example, I’m wondering if “data is contextual” and “data has value” are truly independent concepts) and trying to figure out if there are more concepts in this field.

I would love to hear other people’s ideas about threshold concepts in data management. Has anyone had an “aha!” moment about something that really affected the way they think about data management? Let me know!

Posted in dataManagement | Leave a comment

Data Dispute Prompts Lawsuit

I talk a lot on this blog about one of my big personal interests, data management, but I’m always excited to have an excuse to discuss another interest of mine, university data policy. Today’s excuse to delve into policy comes from one of my data-policy-research collaborators, who sent me a data story so thorny that I just had to discuss it here on the blog.

The case involves a prominent Alzheimer’s researcher, Paul Aisen, who ran the Alzheimer’s Disease Cooperative Study at UC San Diego and just took a job at new Alzheimer’s center run by USC. Aisen is taking 8 staff members with him to the new center, plus his National Institute on Aging grant and its corresponding data. Unfortunately, UC San Diego says that Aisen does not have permission to transfer these grant resources to USC. The data is particularly sticky issue here, as UC San Diego is alleging that the researcher transferred the data to an Amazon server and won’t share the password with UC San Diego administrators. The result is that UC San Diego – or more specifically, the UC System Regents – are now suing both Aisen and USC over the money and the data.

There’s a few issues going on in this case that are worth discussing. First, can the researcher take the grant to another institution? Second, who owns the data? Third, can the researcher take the data to another institution?

The first issue involves grant administration. The news article about this lawsuit states that “university declined to let [Aisen] keep the associated government funding.” UC San Diego likely has some authority to do this as grants are usually given to universities to administer on behalf of the researcher and not directly to the researchers themselves. So while most institutions allow researchers to transfer grants when they move jobs, it’s not necessarily a given – especially where funding covers a whole center rather than a single research group.

The second issue is actually the clearest of the three. University of California System policy states that all data generated by research on campus is owned by the university, or more specifically the University Regents (who are the official suing party in the lawsuit). So Aisen does not own this data, the University Regents do.

However, just because the university owns the data doesn’t mean a researcher doesn’t have rights to the data when he/she leaves the university. PI’s at UC schools are allowed to take a copy of the data with them but can’t take the original without written permission from their Vice Chancellor for Research (this presumes that the data is not “tangible research material”, which the researcher cannot remove at all without written permission). So at the very least, university system policy states that Aisen cannot prevent UC San Diego from accessing and maintaining the master copy of the data. On the flip side, Aisen should be able to take some data with him to USC but it would only be a copy of the data for which he was listed as PI on the grant and not the whole study dataset, which dates back to 1991.

So without knowing the specifics of the case, I would say that UC San Diego seems to have a good claim to the data. This directly results from having a clear data policy.

My own research has found that such university data policies are becoming more common but are far from ubiquitous. While these policies do provide important clarity, anecdotal evidence – like this story – suggests that universities are mainly leveraging these policies when significant amounts of money or prestige are involved. I think that’s a shame because such policies can be very helpful for data decision making.

The other key issue here is the fact that the university owns the research data. This is something that many researchers are uncomfortable with but is often a routine part of doing research at a university; it’s akin to the university claiming patent rights. That said, individual researchers usually get to make most all decisions about the data (in their capacity as data stewards) and should expect something in return for this deal. Namely, universities should take their ownership claim seriously and devote enough university resources to the care and maintenance of “their” data.

I’m looking forward to hearing more details about the case and going beyond my personal speculation to see how things are resolved. In the meantime, it’s makes for another good story to share on the importance of clear data policy.

August 2015 addendum: It looks like my initial assumptions that UCSD owned the data were correct. This LA Times article details how a judge ordered the data returned to UCSD.

Posted in dataDispute, ownership | Leave a comment

Data Management Videos

I’ve been so busy talking about documentation on the blog recently that I’ve forgotten to share an awesome project that I’ve been working on: the data management video series!

Over the course of the last semester, I worked with an intern to create a series of 10 data management videos. The videos cover a range of topics and are all available on YouTube, so not only can you watch them whenever but you are also free to embed them on other webpages. I’m all for sharing content and, while these videos were predominantly made for researchers at my university, the more researchers who learn this stuff the better.

The full series list is as follows:

(As a geeky aside, I also want to point out that I’m wearing some of my favorite handmade items in a few of those videos. Keep an eye out for the epic sweater of awesome, the bad passwords dress, and the marvelous woman-in-science dress as you watch!)

These 10 videos are a solid start to work in this medium and I’m hoping that we can add more to this series over time!

Posted in dataManagement, video | Leave a comment

Taking Better Notes

I’ve been talking a lot about documentation on this blog over the last few months but there is definitely one more issue I need to address before we move onto other topics: taking better notes. Taking better notes is really at the heart of improving your documentation because this is the main way that researchers document their work.

To review, having sufficient documentation is central to making your data usable and reusable. If you don’t write things down, you’re likely to forget important details over time and not be able to interpret a dataset. This is most apparent for data that needs to be used a year or more after collection, but can also impact the usability of data you acquired last week. In short, you need to know the context of your research data – such as sample information, protocol used, collection method, etc. – in order to use it properly.

All of this context starts with the information you record while collecting data. And for most researchers, this means taking better notes.

Most scientists learn to take good notes in school, but it’s always worth having a refresher on this important skill. Good research notes are following:

  • Clear and concise
  • Legible
  • Well organized
  • Easy to follow
  • Reproducible by someone “skilled in the art”
  • Transparent

Basically, someone should be able pick up your notes and be able to tell what you did without asking you for more information.

The problem a lot of people run into is not recording enough information. If you read laboratory notebook guidelines (which were established to help prove patents), they actually say that you should record any and all information relating to you research in your notebook. That includes research ideas, data, when and where you spoke about your research, references to the literature, etc. The more you record in your notebook, the easier it is to follow your train of thought.

I would also recommend employing headers, tables, and any other tool that helps you avoid having a solid block of text. These methods can not only help you better organize your information, but make it easier for you to scan through everything later. And don’t forget to record the units on any measurements!

Overall, there is no silver bullet to make you notes better. Rather, you should focus on taking thorough notes and practice good note taking skills. It also helps to have another person look over your notes and give you feedback for clarity. Use whatever methods work best for you so long as you are taking complete notes.

Research notebooks have been used for hundreds of years. We can still refer to Michael Faraday’s meticulous notes or read Charles Darwin’s observations that lead to the theory of evolution. These documents show that handwritten research notes have been and will continue to be useful. But to get the most out of your research notes, you need to start by taking better notes.

I challenge you this month to think about your research notes and work to take clearer, more consistent, and more thorough notes. Your ultimate goal is to make sure you have all of the documentation you need for whenever you use your data.

Posted in documentation, labNotebooks | Leave a comment

Data or It Didn’t Happen

There’s a story in the news this week about the requested retraction of a study on changing people’s minds on same sex marriage. I always find it interesting when retraction stories are picked up by major news outlets, especially when the article’s data (or lack thereof) is central to the reasons for the retraction.

The likely retraction (currently an expression of concern) in question concerns a study published in Science last year looking at the effect of canvassing on changing people’s minds. Study participants took pre- and post-canvassing online surveys to judge the effect of canvassing on changing opinions. While the canvassing data appears to be real, it looks like the study’s first author, Michael LaCour, made up data for the online surveys.

The fact of the faked data is remarkable enough, but what particularly interests me is how it was discovered. Two graduate students at UC-Berkeley, David Broockman and Joshua Kalla, were interested in extending the study but had trouble reproducing the original study’s high response rate. Upon contacting the agency who supposedly conducted the surveys, they were told that the agency did not actually run or have knowledge of the pre- and post-tests. Evidence of misconduct mounted when Broockman and Kalla were able to access the original data from another researcher who posted it in compliance with a journal’s open data policy. They found anomalies once they started digging into the data.

In my work, I talk a lot about the Reinhart and Rogoff debacle from two years ago where a researcher gaining access to the article’s data led to the fall of one of the central papers supporting economic austerity practices. We’re seeing a similar result here with the LaCour study. But in this case, problems arose due to a common practice in research: using someone else’s study as a starting point for your own study. Building from previous work is a central part of research and bad studies have problematic downstream effects. Unfortunately, such studies aren’t easy to spot without digging into the data, which often isn’t available.

There’s an expression that goes “pictures or it didn’t happen,” suggesting that an event didn’t actually take place unless there is photographic proof. I think this expression needs to be coopted for research to be “data or it didn’t happen.”  Unless you can show me the data, how do I know that you actually did the research and did it correctly?

I’m not saying that all research is bad, just that we need regular access to data if we’re going to be able to do research well. We can’t build a house on a shaky foundation and without examining the foundation (data) in more detail, how will we find the problems or build the house well?

So next time you publish an article, share the data that support that article. Because remember, data or it didn’t happen.

Posted in openData, researchMisconduct | Leave a comment

Templates

In my last post, I discussed my philosophy on documentation in that most researchers need to take better notes and augment them with a few key types of documentation, as needed. I’ve already blogged about a few of these special documentation types – data dictionaries, README.txt files, and e-lab notebooks – but one structure we haven’t examined here is templates. Let’s correct that now.

Templates are one of my favorite recommendations for adding structure to research notes and making sure that you’ve recorded all of the necessary information. They coopt the benefits of a formal metadata schema – making documentation easy to search across, helping you record all essential information, and providing consistency – without all of the fiddliness or rigidity. This makes templates much easier to adopt and use.

So how do templates work? Basically, you sit down at the start of data collection and make a list of all the information that you have to record each time you acquire a particular dataset. Then you use this as a checklist whenever you collect that type of data. That’s it.

You can use templates as a worksheet or just keep a print out by your computer or in the front of your research notebook, whatever works best for you. Basically, you just want to have the template around to remind you of what to record about your data.

Let’s look at an example. When I was a practicing chemist, there were a few critical pieces of information I needed to record every time I ran an experiment. This list included the following:

  • Date
  • Experiment
  • Scan number
  • Laser beam powers
  • Laser beam wavelengths
  • Sample concentration
  • Calibration factors, like timing and beam size

Using this list as a template, I would then record the necessary information every time I did an experiment. The result might look something like the following:

  • 2010-06-05
  • UV pump/visible probe transient absorption spectroscopy
  • Scan #3
  • 5 mW UV, visible beam is too weak to measure accurately
  • 266 nm UV, ~400-1000 nm visible
  • 5 mMol trans-stilbene in hexane
  • UV beam is 4 microns, visible beam is 3 microns

Basically, the list is memory aid to make sure my notes include everything they should for any given experiment. And I could even use different templates for different types of experiments to be more thorough.

Remembering to record the necessary details is the biggest benefit of using a template, as this is an easy mistake to make in documentation. Templates can also help you sort through handwritten notes if you always put the same information in the same place on a notebook page. Basically, templates are a way to add consistency to often chaotic research notes.

I challenge you to try out a template or two and see if they help you record the better notes. Because, as I’ve said before, research data without documentation are useless and, honestly, having insufficient documentation can be just as frustrating. So make your data better by using a template!

Posted in documentation | Leave a comment