accomplishments and lessons learned from 6 years of program development - PowerPoint PPT Presentation

1 / 35
About This Presentation
Title:

accomplishments and lessons learned from 6 years of program development

Description:

National Park Service, Prairie Cluster Long-term Ecological Monitoring Program ... park level. the ability to integrate within the program, region and ... – PowerPoint PPT presentation

Number of Views:43
Avg rating:3.0/5.0
Slides: 36
Provided by: Mik7601
Category:

less

Transcript and Presenter's Notes

Title: accomplishments and lessons learned from 6 years of program development


1
Data Management in the Prairie Cluster Prototype
Long-Term Ecological Monitoring Program
accomplishments and lessons learned from 6 years
of program development
  • The role of data management in a long-term
    monitoring program (LTEM)
  • Data management and the organizational structure
    of an LTEM program
  • Database development - planning for data
    integration and exploration

2
Contributors
Brian Witcher - data manager / GIS
specialist John Boetsch - plant ecologist Mike
DeBacker - botanist
3
Data Management Is More Than A Step in the Process
data collection
data processing
data reporting
data management
4
Data Management Is Integral to the Process
raw data
working data file
reliable working file
archived copies of field sheets
data acquired from third parties
periodic trend analysis
routine summary reports
master dataset
edit log
user interface and integrated database

publicly available dataset
Prairie Cluster Prototype, Data Management Plan
5
Role of Data Management Prior to Data Collection
  • design thorough field sheets that promote
    complete data collection
  • use standardized habitat descriptors and
    abundance classes where possible
  • assure QA/QC procedures are incorporated into
    field methods

6
Role of Data Management During Data Collection
  • Document in a standardized trip report format
  • deviations from the protocol
  • field conditions
  • unusual or unexpected circumstances that may
    affect data quality
  • anything that may help explain trends in the data

7
Role of Data Management During Data Processing
  • database design and development
  • create data entry forms with QA/QC features to
    prevent transcription errors when converting data
    into digital format
  • archive duplicate copies of field data sheets

8
Role of Data Management During Data Analysis and
Reporting
  • automate summary reports
  • promote distribution of results (e.g. websites)

9
Role of Data Management During Data Archival
  • maintain a secure computer environment for data
    archival
  • store duplicate copies of archived data
  • maintain an edit log for archived datasets
  • coordinate population of service-wide datasets
  • respond to data requests

10
Data Management and Geographic Information Systems
Data management plays an important role in a GIS
program. - and - Ecological data must have a
spatial component. - however - Data management
is not just managing spatial data.
11
Dont Let Network Administration Strangle Your
Data Management Program
  • Data Managers are responsible for
  • data redundancy, back-up and off-site archival
  • organization of archived files
  • data security through permissions and
    privileges

12
Be Strategic Regarding the Focus of Data
Management Efforts
Focusing data management efforts early in the
data collection process will produce the greatest
return in terms of data quality.
The hypothetical cost of assuring good data
quality at various phases
high
Cost (effort)
Good Data Curve
low
collection
handling
reporting
archiving
Phase (time)
Prairie Cluster Prototype, Data Management Plan
13
Be Strategic When Focusing Data Management Efforts
  • Take advantage of existing resources
  • WASO database tools (NPSpecies, NPBib, Dataset
    Catalog, etc.)
  • WASO database template, and Data Management Plan
    template
  • prototype program Data Management Plans and
    database designs
  • Define your programs unique features and talents
  • Recognize that many aspects of the NPS data
    management system are in development and everyone
    has an opportunity to contribute

14
  • The role of data management in a long-term
    monitoring program (LTEM)
  • Data management and the organizational structure
    of an LTEM program
  • Database development - planning for data
    integration and exploration

15
Responsibilities of Staff in a Collaborative Data
Management Model
Data Manager Program Coordinator Ecologi
st Botanist Bio-technician
assure overall data QA/QC data archiving and
dissemination database development, report
automation
analysis, data summary and reporting
data validation, summary, reporting and
analysis oversee data entry and verification
data entry, verification
data entry
16
Advantages of a Collaborative Data Management
Model
  • synergy from interaction among individuals with
    different skills and perspectives
  • smoother transitions when personnel change
  • broader awareness, understanding and
    appreciation for data management

17
Prairie Cluster Staff Resources Dedicated Towards
Data Management
Prairie Cluster Prototype, Data Management Plan
18
  • The role of data management in a long-term
    monitoring program (LTEM)
  • Data management and the organizational structure
    of an LTEM program
  • Database development - planning for data
    integration and exploration

19
Key Elements of the Database Development
Process examples from the Prairie Cluster
  • Databases developed incrementally as modular
    products
  • Databases developed by Prairie Cluster staff
    using MS Access
  • Database queries, macros and property settings
    are generally favored over SQL and VB programming
    - code is used judiciously to handle things not
    easily done otherwise
  • Documentation is critical - whether code or
    macros, database designs are documented using
    clear, understandable language

20
Key Elements of the Database System
Design examples from the Prairie Cluster
  • Databases include standardized core metadata
    tables describing the time and place of data
    collection to facilitate data integration
  • Databases have well-developed data entry forms
    to prevent transcription errors and validate data
  • Modular databases refer to shared lookup tables
    that are housed centrally to encourage
    compatibility among data sets

21
The Data Management Mantra
Garbage In, Garbage Out
22
Data Entry Forms Balancing Design Complexity
and User Proficiency
  • designers skill level
  • users skill level
  • stage of protocol development
  • stage of database development
  • complexity of the data

23
A Sequenced Approach to Data Entry
Sample date
Sample conditions
Tree data
Primary habitat data
Vegetation structure
24
A Unified Data Entry Form
  • consolidates data entry for many tables into a
    single form
  • unbound forms provide flexibility to mimic field
    data sheets
  • facilitates efficient entry of multiple records
  • allows users to review the entire suite of data
    before submission

25
What Are the Information Needs
How are changes in grazing practices affecting
prairie health?
plant community
bird community
aquatic community
26
The Value of Data Integration
  • Trends that are corroborated by several data
    sources increases the body of evidence upon which
    to make management decisions
  • Integration promotes regional and
    interdisciplinary data exploration and analysis

27
Some Alternative Mechanisms for Data Integration
  • Integration objectives An approach to data
    integration must balance
  • functionality at the park level
  • the ability to integrate within the program,
    region and service
  • Data integration methods attempted in the Prairie
    Cluster Program
  • a centralized database approach
  • a federated database model

28
Difficulties Encountered with the Centralized
Approach
  • The format and composition of core tables is
    inflexible producing a proliferation of tables to
    accommodate the specific needs of each monitoring
    project
  • It is difficult to anticipate the design
    requirements for future databases when deciding
    on the architecture of core tables
  • As we began to centralize our modular datasets,
    the intractability of the resulting database
    became apparent

29
Difficulties Encountered with the Centralized
Approach
30
Graphic Representation of a Federated Database
archived data sets
user input
grassland birds
graphic user interface (GUI)
retrieve summarized or raw data on demand
grassland plant communities
adjacent land use
integrated data set
aquatic macroinvertebrates
export to regional or nationwide databases
integrative analysis
31
Data Integration
Data dumped when finished (avoids version
problems)
32
Advantages of a Federated Database System Design
  • allows greater flexibility in accommodating the
    needs of each project area
  • allows individual databases to be modified
    without affecting the functionality of other
    project databases
  • avoids the large initial investment in a
    centralized database and the concomitant
    difficulties of integrating among project areas
    with very different structural requirements

33
Plan on Change
  • Change is a sign of a healthy, active data
    management program - not a sign of defeat
  • Dont over-invest yourself in any particular
    design or model, always try to consider new
    alternatives
  • Refinement is never ending

34
Beware of the Mistakes Weve Made
  • Standardize lookup tables from the onset -
    particularly species names
  • Agree on standardize field formats early, pay
    attention to every detail
  • Plan to accommodate different temporal scales
    within and among monitoring projects
  • Determine reference frame and unit of
    replication prior to automating reports

35
What is Memorable About the Prairie Cluster?
  • Recognize the prominence of data management in a
    monitoring program
  • Data management is a task best accomplished by a
    team
  • Given the rate of technical and programmatic
    improvements, a data management program will be
    continually refined
  • Build on the contributions of others
Write a Comment
User Comments (0)
About PowerShow.com