How long and when do you transcribe records in Notes from Nature, and other neat ways to look at your (amazing) effort.
This post is co-written by Julie Allen and Rob Guralnick
Last post, we provided a look at cumulative transcription efforts across our four current projects. These summary numbers hide all kinds of interesting details about how these transcriptions take place and where the data is coming from. Thanks to efforts on the part of guest bloggers Julie Allen (http://wwx.inhs.illinois.edu/directory/show/juliema) and Libby Ellwood (https://www.idigbio.org/content/welcome-libby-ellwood-new-postdoctoral-scholar-idigbio), we have some more in depth statistics to show you.
First, where has your effort helped us get a better understanding of biodiversity? The map below shows you a count of number of records transcribed per country across the world. Not surprisingly there has been a huge number of transcribed records that come from the United States (where we have 407,575 transcriptions completed). However, we now have transcriptions from 175 countries! Costa Rica, for example, has 6,766 records transcribed, 28 in Thailand and 26 in Mali. Explore more in the map below:
We have also been asking some questions about transcription effort and time put in by all the amazing people who have worked on Notes from Nature. So, simple questions: When do people transcribe records? And how long does it take? In terms of the “when”, here is a graph of number of transcriptions and when they happen during the day for the Herbarium collection (based on GMT time). Although its hard to know what time zone our transcribers might be in, we see a lot of activity at all hours (even and maybe especially late night) We wish we could provide more coffee for those working late.
Not all records are created equal in terms of time needed to transcribe them, and surprisingly not all collections are equal either! We know that some records are harder to transcribe than others but on average it takes 3:05 min for a Herbarium record to be transcribed whereas only 2:16 for Macrofungi and the fastest of all 2:04 for an average CalBug record. There is a lot of variation around those times as you can see from the plot below.
Amazingly, there have been 20,761 people helping to transcribe records across the four collections! We find a classic pattern in citizen science transcription projects where the majority of transcriptions have been performed by just a few people, while most folks only contribute briefly (a record or two) and then move on. We’d really like to find a way to engage folks for the long term and change that equation, but as you can see from the graph below, we also have a similar result with a few people doing the lion share of the work, with one person logging an amazing 92,528 transcriptions!!
We hope these summaries provide some useful information. We learn a lot from these, all with the goal of hopefully improving how well Notes from Nature can work in the future. We have some neat ideas about those improvements and hopefully will be sharing some of those improvements just as soon as they are ready. If you want to see the code that generated these results, we’ve posted it to github (https://github.com/juliema/NfN_DataParsing) and will make some of the results data available soon too.