Auditing serves as a set of tools and guidelines to assist in the quality control of ARU projects. Organization and project administrators have the ability to edit any tag data produced in a project, such as modifying species identifications, tag metadata or task metadata. The goal is to have the ability to detect and correct species identifications error in order to increase data quality.
Note: Check with your project administrator for how you should audit acoustic data.
Data Discover now provides a quick way of searching for particular species in a project.
- Go to Data Discover and click the Species tab
- Select the ARU sensor
- Select any projects you want to include
- Select a date range (if applicable)
Clicking on the beside the project name will reveal the task metadata associated with the species detection. The nested tables in the Results section contain the project and task where every tag was detected. You can click on any of the tasks as long as you're a project member. You must be a project administrator in order to edit tags.
5.5.1 Conducting an audit
In the acoustic tagging page, the audit tab is what tracks the changes made in the task. When a project administrator or any user other than the assigned tagger makes changes to the task metadata or tags, it is visible in the audit table. The pertinent database fields include:
- Tag creation date
- Tag deletion date
- Whether the tag is active
- The user who created the tag
- The user who deleted the tag
Here are some types of auditing you may proceed with:
- Full task audit
- Task metadata is checked
- All tag metadata are checked (biotic and abiotic)
- All species identifications are checked
- Species identification audit (biotic audit)
- Confirming, denying or changing specific species identifications in a task
- Erroneous species audit
- Searching and checking species that are either out of spatial or temporal range
- Abiotic audit
- Confirming, denying or changing only abiotic data tag and task metadata
- Go to Data Discover and select the Species tab and the applicable project.
- In the Results table, expand the Project drop-down menu to view all the tasks
- Conduct an audit for each species for the first 15 tasks in the list - if there are less than 15 tasks conduct an audit with however many tasks currently appear
- When auditing a tag, conduct the following procedure:
- If you agree on the identification of the species, the individual ID, the vocalization, the abundance and the confidence, change the Confidence field to Confirmed
- If you agree the tag is a detection but cannot identify the species:
- Change the species to an Unknown category (UN*) with as much precision as possible (e.g. UNPA if a passerine) and change the Confidence field to Changed to Unknown, or preserve the species code name and change the Confidence field to To Be Checked
- If you agree the tag is a detection but cannot identify the species because it is too faint or too far away, change the species to an Unknown category with as much precision as possible and change the Confidence field to Unknown Far
- If you disagree with the identification of the tag:
- Change the Species Code, Individual ID and Vocalization Type where applicable OR
- Change the Confidence code to To Be Checked OR
- Change the Confidence code to Check Data
- When all the necessary changes are made for the task, switch the task status to Review to flag to the assigned tagger that you have made an audit. If no major changes to the task have been flagged (i.e. misidentifications, etc), the task can be kept as Transcribed whereby the original user will not need to make any more changes or modifications.
5.5.2 Troubleshooting auditing
- If the task status is greyed out, it is because the project is likely in a Published status. The project status can be modified in the project settings.
- If you receive an individual ID conflict or constraint error, it is because the system is still being designed to track both active and inactive tags and any conflicts will still arise. For example, this will usually happen if you delete an OVEN 1 tag and then try to repopulate the same individual.
- Certain projects, tasks and tags were imported from an old database where we used the time of first detection and the average frequency range and time-span of a vocalization to populate tags. Tags where the identification is correct but the position or size of the tag will need to be adjusted.