Surprising similarities between field biology notetaking and micro notetaking

Posted by Rory on October 26th, 2011 @ 4:48 pm

Two radically different kinds of note taking?

This week I came across ButterflyNet — a mobile capture and access system for field biology research developed by the Stanford HCI group.  I discovered ButterflyNet in a footnote to Finders/Keepers:  a Longitudinal Study of People Managing Information Scraps in a Micro-note Tool on a ‘micro-note taking tool by David Karger and others at MIT and the University of Southampton.  These are two examples of note taking at the extreme poles of the spectrum: notetaking in the field is complex, with entry of heterogeneous kinds of data, over extended periods of time, often in difficult conditions, whereas micro-notes are scraps of information — like  ‘to do’ items and web urls — that people want to note down as simply and as quickly as possible.  As you’d expect, there are some differences between what note takers want from tools in the two different situations, but what I found more interesting was the number of underlying similarities.

Differences

Field biology

Field biology is a complex activity.  It is carried out in extended sessions, involves observations of multiple things and activities, some of which are changing while the observation takes place, often involve taking photographs and/or collecting physical samples, and requiring recording location information about the subjects of the study.  This results in collection and production of often large amounts of heterogeneous forms of data.  The collection of data in the field, moreover, is only one step in an extended research process.  After the data has been collected in the field, it will be organized, analyzed and often tested back in the lab, with results subsequently discussed and reported in meetings, presentations, and publications.

The paper presenting the results from a trial of ButterflyNet points out that this results in some key requirements:

  1. Capturing and accessing heterogeneous data
  2. Transforming and integrating this data
  3. Robust tools

Micro notetaking

Micro notetaking refers to taking quick notes or recording scraps of information. The kind of thing you put down on a Post-it note or its many electronic wannabe equivalents.  The Finders/Keepers authors found that people wanting to take quick electronic notes have three requirements for notetaking tools; the tools must be:

  1. Easy to use.
  2. Organized to let people arbitrarily capture small bits of information quickly and easily.
  3. Able to keep these information items readily available in visible locations.

Similarities

So much for the differences in the two note taking environments, and the resulting differences in the requirements of field biologists and micro notetakers.  But underlying these differences are some fundamental similarities.  These are reflected in the following feedback from the biologists involved in trialling ButterflyNet:

  1. The in-field focus — when time is expensive — is on documentation, rather than interface manipulation.
  2. The top advantage of ButterflyNet was that it would help participants to capture and transcribe more data
  3. Participants would rather save field time, even if it resulted in more work later.

So it turns out that for field biologists, just like micro note takers, the primary concern is to make data entry as quick, simple and easy as possible, even if this results in more work later on.

Consequences for tool design

These conclusions lead to some important, if difficult, consequences for those of us designing tools for biologists.  The main ones, it seems to me, are:

  • Data entry must be ultra simple and easy — a major challenge given the heterogeneity of the data being entered.
  • This inevitably increases the complexity of dealing with and organizing the data after it has been entered, so it’s a double complexity — heterogeneous data with little structure captured at the time of entry.
  • Users still want the ability to structure data for the purpose of  analyzing and testing it in the lab, and subsequently discussing and reporting data and results in meetings, presentations, and publications.
  • The second challenge is thus building support for organizing and dealing with the data in ways which (a) appear simple to the user, but (b) allow the user to organize, manipulate and present the data, and integrate it with other data (from databases, the web, etc.) in a variety of sophisticated ways.

What kind of field biology notetaking tool would this mean in practice?

ButterlyNet was produced in 2005.  It’s safe to describe it as (a) a brave attempt using the technology then available, and (b) way ahead of its time.  The main curiousity, from the perspective of 2011, is ButterflyNet’s creative attempts to integrate paper note taking with online data capture and retrieval.  This was highly innovative in 2005, but in the age of the iPad, it makes ButterflyNet look extremely clunky.  The question that leaps to mind today is:  why not just get rid of the paper and do it all (a) online and (b) on a tablet?

What would that take?  A tool that

  1. (Like Butterfly Net) Supports entry of heterogeneous kinds of data including notes, photos, audio, video, GPS data, and data relating to physical samples.
  2. Allows simple and quick entry of all these kinds of data, on tablets and offline if necessary (hence dispensing with the need for paper).
  3. Facilitates easy association of the different kinds of data; automatically, or by the user in ways the user determines.
  4. Syncs between tablets and pcs or macs.
  5. Is designed to be collaborative, i.e. has support for controlled sharing of data.

 

If a tool like that sounds interesting to you, watch this space!

Paper lab book versus electronic lab notebook: a real life comparison

Posted by Rory on October 10th, 2011 @ 12:57 pm

An entire paper lab book appended to a PLoS ONE paper!

Gregory I. Lang and David Botstein recently published a paper (A Test of the Coordinated Expression Hypothesis for the Origin and Maintenance of the GAL Cluster in Yeast) which is attracting attention because  the supplementary data consists of a 101 page lab book containing all the notes, strain construction, methods and raw data that went into producing the paper.  As Mark Hahnel pointed out in SciCrunch, this is an admirable example of both open science and thorough science.  But as Mark and several people who commented on Mark’s post noted, the authors would have (a) saved themselves a lot of time and (b) made it easier for themselves and others to make use of the data they generated if they had recorded their data electronically.

To bring this point home, I thought it would be interesting to illustrate it in a very concrete way, by showing how the data from the paper could look, and be integrated with other parts of the research, like protocols, constructs, and strains, when gathered and presented in an electronic lab notebook.  First, here’s how it looks using the paper approach.

The paper lab book view

The paper itself follows a standard format with the following headings:

  • Abstract
  • Introduction
  • Results
  • Discussion
  • Materials and Methods
  • Supporting Information
  • Acknowledgements
  • Author Contributions
  • References

And the paper lab book also follows a familiar format, with dated entries.  The entries contain the narrative of what was done, thoughts about results that were emerging, calculations, formulas, materials and constructs used in the experiment, etc.  Here’s the first entry

 

Benefits of using an electronic lab notebook

Headings mirror publication format

Using an electronic lab notebook would have given the authors a big win right at the outset, because the electronic lab notebook can be set up with the same headings that will be used in the paper where the experimental results are presented and discussed, i.e. Abstract, Introduction, Results, Discussion, Materials and Method, Supporting Information, Acknowledgements, Author Contributions, References.

To illustrate this point, here’s how the blank template looks in the eCAT electronic lab notebook

So  the data for each section of the paper is already organized and available for selective export into the document used for publication.

Presentation stays the same

Here’s what the Notebook field looks like with the initial entry filled in – just like it does in the paper notebook.

Flexible set up

There is plenty of flexibility in formatting the form used to document experiments. In this example it’s been set up so that the fields exactly follow the sections in the PLOS ONE publication, and one additional field called Notebook has been added at the top. That’s for recording the daily narrative as in the paper notebook.  But the form can be set up with any fields  — e.g. as follows where certain fields — Abstract, Introduction, Acknowledgements, etc. — have been removed to simplify the experiment form.

An entire working environment

With the electronic lab notebook the notebook is not isolated from the other parts of the research.  In addition to experiments,  protocols, constructs, etc. can also be created in the notebook.  For example, here’s a protocol

Linking

A big benefit of the electronic lab notebook over paper is the ability to make links.  So for example in the Materials and Methods section of the experiment, a link could be added to the protocol used in the experiment, as shown below

Now, whenever it’s necessary to see the protocol used in the experiment, either when the experiment is active or afterwords, e.g. when it is being written up, just click on the link and you are taken to the protocol. Links can also be made to the constructs used in the experiment, and even the strains. It’s also possible to link to a record showing where the strains or other samples used in the experiment are stored in the freezer, and also what has been done to them in the experiment.

Sharing with lab members

PI and postdoc

The article contains the following note explaining the division of labor between the two authors:

“Conceived and designed the experiments: GIL DB. Performed the experiments: GIL. Analyzed the data: GIL DB. Wrote the paper: GIL DB.”

So the paper notebook is Greg’s notebook.  And, as the note says, he is the one who performed the experiments. Presumably David, Greg’s PI, commented and provided advice as the experiments were being carried out, but this is not reflected in the notebook narrative.

The electronic lab notebook opens up new possibilities for collaboration, and attribution.  For example David could have been given permission to view, and edit, the experiment form, and could use the Discussion field for making his comments as the experiments were carried out, and these would be recorded as part of the experimental record. For example

David could do this at any time, because eCAT is a webapp and hence available 24/7. So he could review the experimental narrative not just when he’s looking at Greg’s paper notebook, and not just when he is in the lab; he could review and make comments from home or when travelling.  So it’s much easier to communicate about the experiment.  And the process is more transparent because there would be a dated record of David’s comments.

Bigger groups

That’s useful enough when  just a postdoc and a PI are involved, as in this case, but it’s even more useful when a larger group is involved in working together on a series of experiments that are going to be written up in a paper.  In that case the group can take full advantage of the electronic lab notebook’s linking and variable permissions capability by, say, having common access to shared resources like protocols and strains, and selective access to the experiment record, e.g. the key contributors have view and edit permission for the experiment record, and others have only view permission.  Links can also be made to the publication in progress. And of course everyone can access all these records in the electronic notebook lab notebook 24/7; the physical and temporal limitations of the paper lab book are left behind.

Sharing with the community

New possibilities are also opened up for sharing with the community.  Whether it is during the course of the experiment —  for open science advocates — or after the experiment has been written up and published, having the experiment documented in the electronic lab notebook makes it much easier to share it with the interested community beyond the lab.  And that sharing has far higher utility, because it allows the authors, and others, to search more effectively.

Search and archiving

It’s just not practical to search for terms, concepts, etc. in a 100+ page paper notebook.  Not even for the author, much less for others like readers of the publication.  Search is useful for those doing the experiment as they carry it out.  E.g. in Greg’s case it would have been useful to search for ‘strain’ + ‘transformation’ + [a particular strain]’.  But these kinds of searches are also invaluable after the fact, both for the researcher and for those wanting to understand, recreate or build on the original research. So with the electronic lab notebook you get a shareable archive of the experimental results and the data associated with them.

Attribution

The electronic lab notebook also keeps an automatic record each time a change is made, so using its audit trial it’s possible to see who made what change when.  This is useful in tracing the evolution of the experiment, and also in determining who contributed and in what way.

iPad electronic lab notebook: the best of both worlds

The paper lab notebook has one major advantage over an electronic lab notebook on a pc or mac– the paper lab notebook is portable.  But electronic lab notebooks that work on mobile devices like the iPad are also portable, so that distinction also becomes a thing of the past.   And the iPad opens up new possibilities for individual mobility and sharing with colleagues.  So you get portability, just like the paper lab book; plus all the benefits noted above of the electronic lab notebook.  It’s the best of both worlds!