beyond counter compliant: ways to assess e-resources reporting tools
TRANSCRIPT
BEYOND COUNTER-COMPLIANTTHE IMPORTANCE OF ASSESSING E-RESOURCES REPORTING TOOLS
University Library
Kelly BlanchatElectronic Resources Support Librarian
Excel Handout: https://tinyurl.com/y7bvlg27
360 COUNTER @ YALE
Out-source usage statistics harvestingTo allocate staff resources elsewhere
Consolidate usage across time and providersTo enhance reporting
NOT to calculate Cost-Per-Use, not using SUSHI
360 COUNTER: HOW IT WORKS
Part 2: INTOTA ASSESSMENT
A. Consolidates usage• Across time • Across providers
B. Assessment (i.e.: CPU)
Part 1: 360 COUNTER
A. Collection of usage statistics• “Data Retrieval Service”• SUSHI
B. Archives usage data
Jan Feb
0 0
1 42
3 7
89 0
Jan Feb
2 4
16 0
0 0
1 0
Jan Feb
2 0
7 6
0 0
26 87
Jan Feb
23 7
0 0
1 1
0 9
Jan Feb
27 11
32 48
4 8
106 96
IMPLEMENTATION TIMELINE
Prior to 2015 Spring / Summer 2015
Administrative credentials added to 360 COUNTER for Data Retrieval Service (DRS)
YUL staff manually uploads historical statistics to 360 COUNTER
YUL staff manually retrieves usage statistics every 6 months
YUL staff uploads statistics to a local website
The DRS team finalizes the first retrieval of Yale’s usage statistics, for January – June 2015
September 2015 November 2015
Collection Development retrieves usage statistics for 2015 OUP BR2 within IA
Oxford English Dictionary is MISSING from consolidated report, as are 342 titles, and 29,170 total uses
October 2015
Begin exploring the consolidated system
Open tickets for data corrections
Pull sample reports from Intota Assessment (IA)
…WHAT DID THAT SAY?
Oxford English Dictionary is MISSINGfrom consolidated report, as are 342 other titles, and 29,170 total uses
November 2015
PHASE 1: TITLE-LEVEL ANALYSIS
Excel V-LOOKUP on ISSN/ISBN and title between COUNTER report and consolidated report
PHASE 1: SAMPLE FINDINGS
360 COUNTER
Missing Titles (20 total titles):• A Dictionary of Geography• The Oxford Dictionary of Plays
Variant Usage (34 total titles):• A Dictionary of Economics• The Oxford Classical Dictionary
Variant ISBN (63 total titles):• Between Two Empires: Race,
History, and Transnationalism in Japanese America
INTOTA ASSESSMENT
Duplicate entries / editions scrubbed
Usage from distinct editions merged
ISBN changed from raw COUNTER
2015 Oxford BR2
REASON
REASON
REASON
PHASE 1: OUTCOME
Title-level example errors submitted to ProQuest for the 2015 Oxford BR2.
After response and fix, only 2 titles and 20 uses remained outstanding.
PHASE 1: SUCCESSFUL, NOT SUSTAINABLE
Favorable results!
Time consuming and really not fun at all
Yale has over 100 providers in 360 COUNTER
PHASE 2: SIMPLIFIED TO COLLECT TOTALS
Collection Date: June 2016
VERY SIMPLE SUBTRACT FORMULA BUILT-IN
PHASE 2: IN PRACTICE
CONSTANT: REPORTSVARIABLE: TIME
RESULT: ACCURACY, OVER TIME
High-level data collection from 8 content providers for JR1 & BR2 submitted to ProQuest’s 360 COUNTER team.
PHASE 2: OUTCOME
360 COUNTER INTOTA ASSESSMENT
HOW
Received granular information on HOW data is consolidated through normalization to the Authority Title.
Jan Feb
0 0
1 42
3 7
89 0
Jan Feb
2 4
16 0
0 0
1 0
Jan Feb
2 0
7 6
0 0
26 87
Jan Feb
23 7
0 0
1 1
0 9
Jan Feb
27 11
32 48
4 8
106 96
CONSOLIDATION = NORMALIZATION
Normalization to the Authority Title will affect overall title count because…
• When duplicates have matching ISSN and title (i.e.: full match), usage is merged onto 1 entry
• When a title has variant data points (DOI, ISSN) over time, titles may display multiple times
• When duplicate titles have the same ISSN but distinct titles, usage is picked from 1 version
EXAMPLE: SAME ISSN, VARIANT NAMES
2015 Springer JR1: When duplicate titles have the same ISSN butdistinct titles, usage is picked from 1 version
EVALUATION
Yale has so much data. High-level analysis has helped us understand WHAT exactly what happening to our data better behind the scenes.
And… it is complex because it is connected to the knowledgebase (and the knowledgebase is complicated).
1) Do the “rules” for title normalization/consolidation make sense?• How does it affect potential CPU reporting?
2) Which results ….• should trigger a bug fix?
• are a result of incorrect COUNTER data?
• are a result of over title normalization?
EVALUATION
EVALUATION: USE MY TEMPLATE!
https://tinyurl.com/y7bvlg27
WHERE WE ARE IN 2017
Prior to 2015 Spring –Summer 2015
Administrative credentials added to 360 COUNTER for Data Retrieval Service (DRS)
YUL staff manually uploads historical statistics to 360 COUNTER
YUL staff manually retrieves usage statistics every 6 months
YUL staff uploads statistics to a local website
The DRS team finalizes the first retrieval of Yale’s usage statistics, for January – June 2015
OED is MISSING from consolidated report, as are 342 titles, and 29,170 total uses
Sept – Nov 2015 April 2016Dec 2015 –March 2016
Phase 1 title-level analysis comparing raw COUNTER reports to consolidated reports in Intota Assessment
Phase 2 begins with high-level analysis, gathering totals between reports
August 2016
Lots of conference calls with 360 COUNTER
Planning for Phase 3 begins, using ARL statistics as a pilot project; transforming COUNTER as data source in Tableau
Lots more conference calls with 360 COUNTER
January 2017
Pilot with ARL stats is complete, internal testing begins in Access, Tableau
March 2017
Add more stats for ARL providers to Tableau
Discuss more robust data solutions (MySQL, Python)
June 2017
NASIG – HELLO!
The Future….
YOU BET THERE’S A PHASE 3!
August 2016
Planning for Phase 3 begins
Transforming COUNTER reports as data source for Tableau
Yay!
• We’re putting ourselves in charge, removing guess-work and assuming “burden” of accepting all data
• Still leveraging the use of 360 COUNTER’s data retrieval – YAY!
• More easily import ILS $$$ data into Tableau to merge with usage
PHASE 3: “LET COUNTER BE COUNTER”
PHASE 3: COUNTER AS DATA SOURCE
STANDARD COUNTER REPORT, JR1
COUNTER REPORT TRANSFORMED AS A DATA SOURCE WITH EXCEL
TABLEAU PLUG-IN
REMEMBER THIS SPRINGER JOURNAL?
PHASE 3: TABLEAU, PROOF OF CONCEPT
2015 Springer JR1: When duplicate titles have the same ISSN butdistinct titles, usage is picked from 1 version – now fixed!
Problem: We collect data on calendar years, but ARL needs fiscal years…(UGH)
PHASE 3: PILOTING
PHASE 3: PILOT WITH ARL STATS
• It’s self service! We can set-up standard views set-up in Tableau for subject librarians to retrieve
• Move from less manual (Excel, Access) to more automated and robust (Python, SQL)
• Data visualization!
• Self-service for renewals
• NO QUESTIONS about data, built on COUNTER standards
NEXT GOALS…
PREPARING FOR WHAT’S AHEAD
Phases 1-3
SUSHI, COUNTER R5