Using Web Metrics to Analyze UseImpact of Education Digital Libraries

1

• Using Web Metrics to Analyze Use/Impact of
  Education Digital Libraries

Kimberly Lightle, lightle.16_at_osu.edu Director of
Digital Libraries Ohio State University College
of Education and Human Ecology
This work is supported by the National Science
Foundation under Grants No. 0840824 and 0733024.
Any opinions, findings, and conclusions or
recommendations expressed in this material are
those of the author(s) and do not necessarily
reflect the views of the National Science
Foundation.
2
Todays Topics

• Overview of projects
• Web metrics and units of analysis
• How we use metrics data to determine content
  development and delivery mechanisms
• Limitations and possibilities

3
What content are we producing?

• 2 websites
• Beyond Penguins and Polar Bears
• NSDL Middle School Portal
• 3 blogs
• Polar
• Connecting News to the National Standards
• Middle School Math and Science
• e-books
• Interactive applets
• Webinars
• TappedIn chat transcripts
• Digital Stories
• Podcasts

4
What real data are we collecting?

• Survey Data
• General web use
• Specific content
• Expert Reviews
• User Feedback
• User testing (homepage, issue page, article
  format)
• Focus groups (OH, TX, NC, CO)
• Comments from Blog, TappedIn, webinars
• In-links (who is linking to us)
• Comments of guest contributors
• Field Test Teachers and Students
• Pre/Post Questionnaire on student attitudes
  toward science, teacher practice/polar science
  knowledge

5
But were web-based

• Our content has the possibility of being
  discovered by anybody
• We especially dont want to be invisible to our
  main audience
• Work towards being as discoverable as possible

6
Maximize Resource Discovery

• Our content is
• Relevant
• Contains key terms and phrases that people use
  when they search
• Has high quality in-links
• Found in multiple places in the cybersoup
• Search engine optimized
• Google has 50 of the market we do what Google
  asks

7
Web Traffic

• In October 2008 over 100,000 page views
• Thats a lot of data
• to mine!

8
Web Metrics Data

• Referrals (front door or side window)
• Search logs
• Search engine referral logs
• Site search logs
• Number of Pageviews
• Sessions
• Requested Pages
• Click Paths
• Time Frames
• Session Length How sticky are pages?

9
Who are our users?

• Intended audience teachers
• Real audience some direct information through
  surveys
• Infer from
• Traffic sources
• Keywords
• Landing Pages
• Top Content
• Comments on blog

10
What are they doing? Where are our users?

• Searching
• Reading blogs
• Editing wikis
• Looking for images
• Being social
• Bookmarking sites
• Sharing bookmarks

11
Referring Pages

• Google
• nsdl.org
• Diigo.com
• Twitter
• Sciencegeekgirl.wordpress.com
• delicious.com
• leecountyschools.wikispaces.com
• nsf.gov
• internet4classrooms.org

12
Where on the web are teachers?

• Digital stories uploaded to YouTube and
  TeacherTube

Only includes clicks on that particular post.
Does not represent all views (general blog
traffic and RSS subscribers).
13
Where are our potential users?

• Putting content multiple places
• Curriki
• NSDL
• ePals
• TappedIn
• NMSA
• Reuse resources across projects
• Where else?

14
What are their needs?

• Site search logs
• Referral keyword logs

15
Keyword Searches(October 2008)

• Water cycle lesson plans
• Discrepant events
• Teaching context clues
• Beyond penguins and polar bears
• Rocks and minerals lesson
• Teaching day and night
• Cause and effect activities
• Expository articles
• Science notebooks
• Seasons misconceptions

16
Top Landing Pages (October 2008)

• Rock and Mineral lessons
• Home page
• Literacy Content Knowledge Context Clues
• Water Cycle lessons
• Discrepant Events in Elementary Classrooms
• Season Misconceptions
• Aurora art lessons
• Rock and mineral misconceptions

17
Top Content(October 2008)

• Rock and Mineral lessons
• Home page
• Water Cycle lessons
• Literacy Content Knowledge Context Clues
• September Issue
• Season misconceptions
• Rocks and Minerals feature story
• Discrepant Events in Elementary Classrooms
• Weather and Climate lessons
• Day/Night and Seasons lessons
• Aurora art lessons
• Rock and Mineral misconceptions

18
Traffic Sources Beyond Penguins

• Decisions
• Maximize search engine optimization
• Solicit links from related sites
• Promote site to increase direct traffic
• Design site to encourage further exploration

• Direct Traffic 6
• Referring Sites 13
• Search Engines 81

19
Traffic Sources Referring Sites

• Referring Sites
• Expert Voices
• internet4classrooms.com
• NSDL
• Byrd Polar Research Center
• NSF
• Yukon Education Student Network
• International Polar Year
• Compare to site averages

• Decisions
• Identify where teachers are on web and solicit
  links/distribute content

20
Multiple Sources of Data from Multiple Sources

• Tell us something about
• Our users
• Their needs
• Where they are on the web
• Quality of resources
• Impact of resources
• And in turn informs
• Content Development Decisions (What)
• Delivery Decisions (How)
• Placement Decisions (Where)

OSU, NSDL, Content Clips, Google
21
Conclusions

• Web metrics are useful, but dont provide
  conclusive data about audience and needs
• Need to triangulate data across multiple data
  streams
• Need to consider context project design and
  users themselves