Monday, April 14, 2014

Computers in Libraries 2014 - Day 2 - What Does the Dashboard Tell Us?

Amy Deschenes of Simmons College gave this presentation on the implementation of a statistics dashboard for the college library's collection and display of statistics.

The library had been storing all of its statistics in a shared Excel file.  This was cumbersome and not particularly efficient.  They decided that they would develop a new system for data collection and reporting.  They went about doing this in three steps.

The first step was to define a data collection process.  To do this they asked their stakeholders for what information was needed to be collected and they determined the different sources of the data, how the data would be sliced up, and how frequently it would be collected.

The second step was to implement the best tool.  They determined that Excel was beyond what they needed for data collection.  They looked at using MySQL and a service called Zoho Creator.  They decided that MySQL would provide too many technical difficulties and went with Zoho Creator's cloud-based database service, which offers a single database in the free version of the product.  They then setup a fixed schedule and had a student input the previous year's data and made sure everyone knew the data input schedule

Now that they had a database that was being regularly updated with the information they created a way to view the data.  This was the third step of creating an online presentation.  They used a JavaScript library called Sheetsee.js which can take the contents of a Google Docs spreadsheet and make a nice interface to filter data and see charts (see their implementation).  They manually copy the data from Zoho each month and put it into the Google Docs spreadsheet for their reporting.

After doing this, they decided to also make a more exciting, interactive, visual and student-focused dashboard.  This they built using a responsive HTML 5 framework called Foundation.  After looking at some other dashboards for ideas they created a friendly dashboard with both quantitative and qualitative data.  The results are quite attractive and friendly.
Although much of their process solved problems that we solved in our own way some time ago, there were approaches here I found useful and I think there are some good things to be learned from their experience.

No comments: