Panel 2: Data to Knowledge

Panel 2- From data to knowledge: the ecoinformatics imperative

Moderator: Jim Reichman
Panelists: Brian Enquist, Matt Jones, Bill Michener, Josh Tewskbury

6 thoughts on “Panel 2: Data to Knowledge

  1. I think that the issue of heterogeneous data quality is key – we have to be able to trust the data without needing to drill down into the details of every dataset. Data collection protocols are one solution, but another might be to develop derived data products that are trusted. My conceptual model for this is satellite data – very few scientists work directly with the raw data stream from the satellite. Instead they work with derived products that try to correct for the vagaries of clouds, view angle, etc… and even further derivations, such as NDVI. Are there analogous, trustworthy, derived datasets for ecology?

  2. NCEAS is a classroom.
    Imagine NCEAS is a classroom. Collectively, we are all the students. Like good educators, the teachers (SAB, Jim, Bill, Frank, Mark) split us up into smaller groups to do our independent projects. We went away, did our work, wrote our term papers, and gave our presentations. However, like a real classroom is there a means for us to have a classroom-level analysis of our work? This symposium right. However, something is missing. I can read the papers and really enjoyed examining the top 100 papers but it would be amazing to see the work that others have done. What if every working group posted a google doc of the best data they collected and their favorite graphic. As a classroom, it would be profound to look across all the files to see patterns and commonalities. It would be also interesting to see how they connect.

    Has one working group ever used the data of another? It would be amazing to see the pamphlets, figures, and data in some form that others have generated through the support of NCEAS. This may also speak to how NCEAS 2.0 could move forward, a library of work done previously (by NCEAS or any other sources) that new groups explore and add to. Inter-working group collaboration. Or the online, ongoing NCEAS science fair for other students, parents (NSF), and the community.

  3. Quick questions for Matt & Bill.
    Is there a critical mass or data density that stimulates effective synthesis? Is there a minimum that we need to begin the process if big picture analyses?

    Do the best syntheses always need quantitative datasets?

    Can synthesis be done on the informatic attributes of datasets and not the data? How valuable or to what extent do ecoinformatics transcend the data? I am just trying to imagine other ways that we do synthesis and get inspired for future activities.

  4. Is the data to informatics ratio too low in ecology?

    Widespread adoption by the greater community including citizen science of any sort of online sharing tool (such as social networks) requires it to be simple and easy. Like most ecologists, I want the gold standard data and ecoinformatics associated with it too. However, there must be examples of simple data, widely distributed with low informatic attributes that could be incredible useful for synthesis. Perhaps we should be asking what is the minimum informatic elements people can provide with their data to make it useful and automate the rest of the meta-data (i.e. read the columns, number of entries, etc) versus shooting for the gold standard.

  5. Following the panel discussion I reflected that there continues to be a need to teach the next generation of ecologists the principles of data management, sound analytical and statistical techniques for synthesis research, as well as to draw them in to the “cozy” culture of collaboration that NCEAS has fostered. Many of us who have been involved in NCEAS take on this responsibility through individual teaching at our respective institutions, but this could be an important area for NCEAS 2.0 to expand and impact a broader audience. Simple, stand-alone curriculum modules for undergraduate ecology lab courses and graduate seminars could be made available via the NCEAS website. These could simultaneously teach principles of ecoinformatics while showcasing key NCEAS data products and publications, and provide a focal point for future grant-writing and fundraising efforts.

  6. I think the idea of NCEAS developing curriculum modules for undergraduate ecology courses is a good idea. I think that much of the success of NCEAS can be attributed to the fact that it came in with a model that had never been applied in ecology before. However, eventually, any model gets stale and if NCEAS is going to continue to make novel contributions, it needs to keep reinventing the model. Development of curricula might be a good direction for such re-inventions, particularly if there might be a funding source for this kind of thing…

Leave a Reply