If the posts about adjustment sets and DAGs have peaked your interest and you would like to learn how (or if you need to learn how - for those taking my pathology class this summer) this post provides links to resources, primarily resources made available by DAGitty.

The first step would be the DAGitty Learn page: http://www.dagitty.net/learn/

The recommended reading at the bottom of that page is excellent and highly recommended, but not required to start making DAGs. The helpfulness of those readings would be for considering the implications of particular causal structures. Since my goal right now is to help get as many physical therapists as willing to be making DAGs based on existing studies and knowledge to sharing, discussing and debating then I think it is fine if you want to jump right in and start making some for sharing. Until there is another platform for sharing I would encourage people to save them on DAGitty and let me  know they want to share it and you can be a contributor to the blog to share your DAG.

The second step (or approach) would be to jump into DAGitty itself and start making DAGs. From the DAGitty home page you can either click on “Launch” or you can download a folder that allows you to launch from your local computer (for when you do not have an internet connection). Note that you can only save your DAG to the DAGitty repository when you are working online, but you can simply transfer the text in the code window from your offline DAG to the online DAGitty workspace and your DAG will then be online and ready to save. But now I am getting ahead of myself.

Here is the home page:

Screen Shot 2015-04-21 at 5.04.30 AM

Click on Learn for a step by step tutorial about DAGs with interactive exercises.

Click on Launch to start making DAGs:

Screen Shot 2015-04-21 at 5.09.56 AM

The “How To” drop down menu is very helpful to get you started making DAGs:

Screen Shot 2015-04-21 at 5.10.53 AM

You use the “Model” drop down menu to create new model, load from, publish on, update from, delete from dagitty.net, or export the graphic itself. If you select choose a new model you immediately need to name your Exposure (can also consider an intervention, cause) and Outcome (effect). If you want to save the code you simply cut and paste the code into a text file.

Screen Shot 2015-04-21 at 5.13.30 AM

To reorganize the layout you can simply click and drag the nodes (variables) around the workspace:

Screen Shot 2015-04-21 at 5.14.50 AM

The “Examples” drop down will provide - well - example DAGs from publications (a great way to learn more about DAGs is to get those papers and compare with the posted DAGs). “Layout” will automatically reorganize the nodes and edges of the graph, and “Help” will open a PDF of the DAGitty manual, which you will be familiar with already from having looked through the DAGitty/Learn page.

In upcoming posts I will continue to provide examples and explanations of DAGs and try to make more explicit connections back to how we got here from where the blog started, how DAGs connect to KBP as a clinical epistemology founded on a critical realist perspective, how DAGs provide an explicit interaction between the empirical and the rationale for developing a knowledge based practice.