Guided Instructions: Correct Records During Quality Control

How to use these instructions:

  • This page lists the specific steps for finding and correcting errors in metadata records

  • Note that each editor only has access to specific parts of the Digital Collections, so not all examples may apply

  • Although this page will include some examples of common errors, editors may need to identify specific types of problems based on knowledge of a particular collection and on relevant documentation:

Step 1: Identify Issues

  • Use the system tools to find problems in metadata, such as mis-formatted or missing values

  • Note that tools will generally not say that something specific is wrong, they just organize information in ways that may help editors identify an issue

  • For each of the tool sections on this page, there is one specific example; use the documentation about the individual tools for more suggestions

Dashboard

  • Some Dashboard filters can identify problems such as “incomplete” records or records containing invalid dates

  • More about Dashboard options

Screenshot of filters in the edit system Dashboard.

Count

  • The count tool helps to find missing or extra entries, such as:

    • Too few subjects (or no UNTL-BS terms in Portal records)

    • Multiple photographers for a single photographer (i.e., possible placeholder values left in a record)

    • No creation date for items in a collection that have known dates

  • More about the count tool

Screenshot of subject entries in the count tool.

Facet

Screenshot of personal creator name entries in the facet tool.

Cluster

  • The cluster tool helps to find values with different formatting or that do not fit a particular expected pattern, e.g.:

    • Near-identical names with different spacing or punctuation

    • Extremely long (or short) subject or title values

    • Identifiers that do not fit a standard pattern used in a collection

  • More about the cluster tool

Screenshot of organizational creator entries in the cluster tool.

Step 2: Find Records

  • If you’re using the Dashboard filters, click on the appropriate filter (e.g., “Invalid Dates”)

  • From the tools, click a value (in cluster or facet) or a numeric value (in count) to open a Dashboard search meeting those criteria

Screenshot showing a cluster value and the related Dashboard search for the value.

Step 3: Edit Records

  • For each record on the Dashboard list for an identified issue, click “Edit” to open the editing form and change, add, or remove the value (depending on what is appropriate)

Screenshot showing record in the metadata editng form and the corrected value.
  • After fixing the issue, click the “publish” button at right to save the record (do not change the visibility of the record unless instructed to do so)

Screenshot of the publish button in a metadata record editing form.

Step 4: Repeat

  • After a particular value has been changed, the lists in the Dashboard and tools will update (note that values in cluster cache and will take longer to reflect changes)

  • Find a new issue or value that needs to be changed and follow the same steps