2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 290-4
Presentation Time: 9:00 AM-6:30 PM


CHANG, Lucy, Department of Integrative Biology and Museum of Paleontology, University of California, Berkeley, Valley Life Sciences Building, Berkeley, CA 94720-4780, luchang@berkeley.edu

Modern studies in biology and paleontology increasingly generate and use large amounts of data. The ability to collect this data quickly and reliably, however, remains a persistent hurdle, and we are in the early stages of using modern technology and global accessibility as research tools to overcome it. As an ongoing large-scale morphometric study of ammonite conch coiling parameters through time, we planned on collecting approximately 5,000 linear measurements. To meet this demand and streamline the data collection process, I constructed a web applet that would allow the user to take constrained measurements of specimen images. The applet takes advantage of widely available Javascript libraries and feeds results directly into Google tools (Forms and Sheets), eliminating user input error and producing data that is easily accessible. This framework is highly adaptable for future projects.

We compared the precision, accuracy, and efficiency of gathering measurements using this web-based approach against measurements of the same figures taken by hand using digital calipers. The precision of measurements taken through the applet was generally on par or better than measurements taken by hand. The difference in mean measurements suggests no consistent bias in the measurements taken by either method. The time spent per figure was also reduced, accelerating data collection. This applet provides a level of transparency that cannot be met by traditional methods. For example, the position and orientation of each measurement were recorded, so that potential outliers could be easily identified and assessed. These results highlight just some of the benefits of integrating web tools into the workflow of data collection. Furthermore, this framework can easily be used for crowd-sourcing data collection. Based on these findings, I encourage the development of tools like this to improve the efficiency and transparency of high-throughput data collection.