Printing Providing and Using Scientific Data Through NSDL


Providing and Using Scientific Data Through NSDL



Cathy Manduca, Carleton College


Ben Domenico, University Corporation for Atmospheric Research


David Halle, University of California at Los Angeles

Andrew Beveridge, Queens College


Susan Van Gundy, University Corporation for Atmospheric Research


New technologies are facilitating unprecedented access to and techniques for interpreting complex datasets, as well as data sharing across disciplines and beyond traditional communities of scientific researchers. Authentic inquiry-driven experiences are being increasingly proven to have positive impact in K-12 and undergraduate education, stimulating a demand for data resources that are engaging and easy to use for non-technical audiences. This session will explore multiple aspects of providing and accessing data through NSDL including technical challenges, evidence from classroom use, new visualization tools, and discussion about how NSDL can best support the differing needs of researchers and educators.

Notes - Providing and Using Scientific Data Through NSDL


Examples of Data Providers


Mapping Application to look at change over time- racial change in New York City.
Underlying material comes from different project- need to get data into shape to allow users to really interact with the data. Need to organize reasonable amount of data in meaningful ways.

Photo Collections:
  1. Camilo Vergara Photo Collection: Multimedia CD

  1. Walking tour of art galleries of Chelsea section in New York

Unidata THREDDS: Making Distributed Datasets more available and usable in NSDL
Future Directions-

Questions:
Who is using this, what level, how any?
Almost entirely at undergraduate level. Universities (80-90) have created on products, based on Unidata technologies, for K-12 environment. One of the main activities of THREDDS is working with DLESE to run workshops- one per year. We bring in technical people, data providers, and content developers and see how we can build modules that can be applicable to middle school environments.

Have to be a Unidata school to use and access?
No. Anyone who is in educational world can work with us. We have hardware grants available. No formal membership is necessary.

How are end users at university level using data?

Work by Cathy Manduca and David Mogk, out of CI sponsored workshop- "Using Data in Undergraduate Science Classrooms"- hard copy available
  1. Want students to work with real world complex problems- excitement
  2. Students need to critically evaluate validity of data-
  1. Use data to illustrate concepts and ideas (pictures in lectures)
  2. Enable student investigations
  1. Have students collect and interpret their own data in context of larger data set
  2. Use existing data to answer questions, often asking new questions
Collect data and develop model, test relationships- workshop physics
  1. Kinds of activities

Repeating Themes-
  1. Finding data hard for non-expert (has to do with indexing)
  2. Use varies with faculty style and course content- there is no ultimate data presentation
  3. Faculty/ students want to be able to adapt and create from data.
  4. Return on learning tool must be worth investment.
Tools that are harder to use are more powerful in what you can do with them.
  1. Importance of critical thinking and evaluating quality of data sets. Providers get concerned about letting out data that isn�t controlled. Should data sets that are questionable be flagged or just not made available?

Discussion: There are some circumstances were you want students to understand how to fill in holes in data, other times you don�t. Teaching data literacy is important in higher education and K-12. There is also a need for highly annotated data sets where you know how the data collection took place. In K-12, there is a need to have data posted with caveats. At undergraduate level, use goes from just wanting a picture to wanting all of the information. Experts have own methods for evaluating data quality- expectations based on where data comes from- once you move out of the expert user, you don�t see that ability to evaluate the data.

Discussion:

No. Using Data study didn't focus on this. Issues come up in issues around tools. Don�t want to have to use different tool for every data set. Want to use one tool. Discussion about how people want to find and combine data sets. Expect data to be combined so didn�t really focus on this.

No comprehensive list right now. But, huge use of Excel.

Earth Exploration Tool Book- K-12 experience, in testing stages

Interoperability Issues, Emerging Protocols Discussion Transcript

Open GIS protocols for accessing data and creating metadata. How do you represent units and geographical location of data sets? Has to be there in some way that you agree on.

Problem not the protocol, but the data.
New York City has done fly-over to create new map/ GIS. Suddenly all census track maps are obsolete. Doesn�t matter which standard they are in, they are wrong. Problem for GIS stuff is the data. There are mapping packages that allow you to translate the data. Protocol/ standard debate irrelevant. If data is not measured the same way, they aren�t going to line up.

Problem neither. People aren�t used to learning from data. We don�t understand how people learn from data. Data doesn�t persuade students. Zero power to overcome the power of a sentence that students read in a textbook. Text carries more weight than data set for students. How do we help people learn form data? How to teach with data is the problem, not what standards to use. Ex: Abstract data into stories.

Abstracting data into stories is problematic because then you are teaching through text, not teaching to read data.
Have students create own stories out of data, don�t tell them the stories.
Practice reading charts and data until comfortable doing so.
Practice is part, not alone
More fundamental than this. Being able to deal with data period. Practice is very important. Don�t do evaluation of data we want them to. With data they have to analyze. Easy to understand and repeat stories/ text.

Analyzing data is a complicated notion. Bring expertise to graph when analyzing data. Need to teach students to bring that expertise or to find out what they don�t know to be able to analyze.

Problem isn�t the data itself, but translating data to text. It's easier to respond to data questions using numbers than translating responses to text. Problems in translation form one mode of though to another.

There is huge power in narrative, but perhaps first step has to be in students creating own narrative. Make own observations and drawing collections.
Questions of how people learn in terms of meta-cognition.

Importance of assessment in overall learning
Way of assessing how students are interpreting data- going back and forth between numbers and text.

Issue of made-up data in K-12 classroom: examples look like data, but are clearly made-up. Why is this so prevalent?
Math community has a problem with this �fake� data. Have to go further than giving students fake data. Have to give students more authentic data so they can translate this to the real world. Make examples �cleaned-up, real data�

Ability to understand a graph is one predictor of success in science courses in undergraduate.
Long path between getting students from point of recognizing data to point of interpreting data.



Comments

Please enter any comments in the following format.