#LawRepositories - Hollie White and Avery Le - Using Metrics to Make Repository Decisions
Collecting the numbers:
The Obvious:
- Number of papers
- Number of downloads
- Email of monthly usage from bepress
- Author reports
Systematic Collecting:
- What’s important to community?
- When will you collect it?
-
- Regularly!
- Who wants to know?
bepress usage reports allow drilling down into the data.
- For DOWNLOADS
Google Analytics.
- For VISITS
Hollie collects:
- Overall and collection level downloads.
- Article level collections
- Visitor information, looking for spikes.
Visits ≠ Downloads ≠ Citations
100 DLs to 1 Citation? Research a bit stale on this.
Communicates raw numbers to library leadership.
Also, to journal editors.
- 9 student-run journals.
Monthly newsletter.
Annual celebration.
What’s Normal?
- Every repository is different.
-
- But, custom alerts in Google Alerts. Close to real-time monitor.
- Spikes in traffic:
-
- Is it a bot?
- Is it conspiracy theorists?
Research Questions
- Correlations and Multiple Regressions for statistical analysis of collection
-
- Analysis in STATA
- Example: Do metadata and downloads correlate?
- Ranking of downloads
Visualizations of metric data:
- Who do you show this to?
-
- Faculty
- Administrators
- Search Queries (from Google Analytics)
-
- Tool: Tagul
- Donut Chart (social media data)
-
- Tool: AmCharts
- Map (network domains)
-
- Tool: Google Fusion Tables
Use — turn metrics into metadata. Some keywords more searched then others? Why not use them?
