First Tuna Data Workshop TDW1

1
First Tuna Data Workshop (TDW-1) 23-27 October
2006, Noumea, New Caledonia
SESSION 5.3 Data Quality Control Systems
Oceanic Fisheries Programme (OFP) Secretariat
of the Pacific Community (SPC)
2
Presentation Outline

• Why Data Quality Control is important
• Quality Control procedures in Data Collection
  systems
• Data collected by Fishing Companies
• Data collected by Fisheries Division staff
• Quality control procedures in Data Management
  Systems
• Pre-data processing
• Data Processing
• Post-data processing
• Data quality reporting
• Identifies problems in data collection systems

3
Why Data Quality is important

• Data Quality is important in ensuring the data
  collected and managed are accurate /
  representative.

• You can have the best database system in the
  world, but if the data collection is not
  appropriate, then summary reports are not
  representative.
• put another way, Garbage in means Garbage
  out

• Data coverage is one component in Data Quality
  Control.

4
Quality control procedures in Data Collection
Systems

• Data collection by Fishing Companies
• Forms to be submitted by Fishing companies should
  be done so as a condition of license (coverage)
• The fishing industry are aware of their
  obligations to provide accurate data
• Training and awareness material for each type of
  data collection are available to fishing
  companies
• There are mechanisms/procedures for liaison with
  fishing companies with respect to (i) the regular
  provision of data and (ii) any problems
  encountered with the data

5
Quality control procedures in Data Collection
Systems

• Data Collected by Fisheries Division Staff
• mostly covers Observers, Port Samplers and Port
  Inspectors
• Mechanisms to ensure the awareness amongst data
  collection team of what others in the team are
  doing
• (including responsibilities)
• There are training courses/resources available
  for data collection staff.
• Comprehensive Data Quality training courses are
  offered for Observers by regional agencies
• A Manual is available for Port Samplers in
  addition to specific in-country training
• There is a clear plan and schedule of first-time
  and ongoing training for data collection staff.
• Forms are the latest versions of the regional
  standard
• Standardisation of data collection

6
Quality control procedures in Data Collection
Systems

• Data Collected by Fisheries Division Staff
  (cont.)
• mostly covers Observers, Port Samplers and Port
  Inspectors
• There is a system to ensure resources to support
  port sampling, observer and other data collection
  activities are always available (e.g. a system to
  keep track of stocks of equipment, forms, etc.)
• There are mechanisms and procedures to (i) check
  the quality of data collected and (ii) provide
  feedback to relevant data collection staff
• Observer (briefing and debriefing systems)
• Port Sampling (Spot checking)
• MCS data (e.g. boarding and inspection)

7
Quality control procedures in Data Management
Systems

• Pre-data Processing
• (The phase when data are being prepared for data
  entry manual checking)
• Batching data
• sorting hard-copy data for easy reference later
• Non-Reporting manual checks (for example)
• Missing vessel details, effort, catch, positional
  information on logsheets
• Mis-Reporting manual checks (for example)
• Catch on logsheets have been clearly recorded in
  the wrong column
• Erroneous dates
• Incorrect totals

8
Quality control procedures in Data Management
Systems

• Data Processing
• (Data Quality Control processes incorporated into
  data entry/verification systems)
• Database system checks (for example)
• Dates of departure from and return to port are
  consistent with fishing operation dates
• Catch (in number and weight) for each species
  falls in the range of acceptable values for that
  fishing gear
• Average weight (weight / number) is an acceptable
  value for that species
• Effort (e.g. hooks set) is in an acceptable range
  for that fishing gear
• Reference table checks (e.g. vessels, ports,
  activity codes, school type codes, species codes)
• Integrity checks (e.g. every trip record has at
  least one fishing operation record)
• Double-entry verification checking
• Control totals checking

9
Quality control procedures in Data Management
Systems

• Post-data entry Processing
• (Data quality control processes incorporated into
  database systems)
• Database system checks not possible during entry
  
• Distance traveled
• Cross-reference to other data types
• Consistency with other trips
• Integrity checks
• Substitution in cases of non-reporting (for
  completeness)
• Catch and effort data using average values

10
Tools for Data Management and Quality Control
(examples of database tools to undertake this
work .)
Procedure to correct logsheet catch with actual
weight of catch
11
Quality control procedures in Data Management
Systems

• Data Quality Control Reporting
• (Reports used to check data quality available in
  TUFMAN)
• Reconciliation with other types of data (for
  example)
• License Fees and Receipts reconciliation
• Logsheet and other types of data have been
  received from vessels that arent in the
  licensing system
• Telex Reports - Logsheets Reconciliation
• Vessel Activity Log - Port Sampling - Unloadings
  Logsheets Reconciliation
• (more to come )
• Checks for erroneous data
• Logsheet and VMS Data - Position Conflicts
• (more to come)

12
Tools for Data Management and Quality Control
(examples of database tools to undertake this
work .)

• Procedures to reconcile the catches and dates
  recorded on
• Vessel activity logs,
• Logsheets, and
• Unloadings

13
Tools for Data Management and Quality Control
(examples of database tools to undertake this
work .)
14
Quality control procedures in Data Management
Systems
5.3 EXERCISE SESSION Group Discussion Are
other Data Quality Control Systems used in your
country that are not included in the list
presented here. Please describe these
systems. (Reconvene and briefly discuss any
additions to be noted and an updated list to be
distributed later)
15
Quality control procedures in Data Management
Systems

• 5.3 EXERCISE SESSION
• Exercise (Individual countries) Using the list
  of Data Quality Control Systems provided,
• Check off each system that already exists in your
  country
• For systems that do not currently exist in your
  country
• Briefly describe why it would be difficult to
  implement, or
• Indicate that it will be implemented, or
• Indicate that it will be reviewed and
  considered
• The output from this exercise will feed into your
  National Procedures Document.