Thursday, November 3, 2011

Toolkit: Atlas.ti

 I thought I'd share some of the tools we're using on this project. When most people think about archaeological tools, they think about tools we use in the field -- shovels, dustpans, brushes, etc. Those help us get our information out of the ground; but then we have to make sense of it.

One of the challenges of the Wiawaka Project is the massive amounts of data that need to (and will need to be) both managed and analyzed. This information includes the layout of the property including building and roadway locations (and former locations); excavated data (what artifacts were found where -- both horizontally across the site and vertically in time); photographs (contemporary and historical); maps (historic, modern, and site plans); primary documents (guest registers, meeting minutes, census data); secondary documents (other writers talking about Wiawaka or other similar sites); background research (similar sites, general historic context, information about the Girls' Friendly Society, information about Troy, information about factory workers at the time, etc. etc.); the list is almost endless, and overwhelming. Not only do we have to keep track of all these various pieces of information, we have to be able to make connections between it all; to make sense of it.

Atlas.ti splash screen, waiting for data to be entered!

Atlas.ti is a qualitative data analysis (qualitative = descriptive data, vs. quantitative = numerical data)  program that allows us to identify themes or threads across many types and formats of document (types = as described above; formats = .pdf, .doc, .jpg, etc. etc.). These are flagged in each document pulled into the program using codes. Once data has been coded in documents (there will be hundreds thousands of documents for the Wiawaka Project; indeed, I have hundreds of site and research photographs already...), Atlas.ti can be used to pull out information matching those codes (themes) from *all* the documents so it can be looked at all together. Relationships between the codes can be made as well, so the program can pull out related themes -- again from across all the documents, or just a subset of them. Making sense of the data comes next, but the program will make it easy to find relevant data.

As a bonus, the program keeps track of where all the digital sources are on my hard-drive. Which makes it easier for me to manage the data storage (rather than building a massive set of relational databases on my own).

I've not used the program before, but will be getting some hands-on practice as part of one of my courses. I am very excited by the organizational solutions the program provides, and the way it will allow me to pull at different threads in the data to help make sense of it.

No comments: