Hi Dedoosers, today we are bringing you an article expanding on a pretty crazy feature added in Dedoose 8 that you may not be making the most of. We are talking about creating Data Set filters via the Excerpts tab! Filtering via the Excerpts tab gives you the ability to exclude or include specific excerpts and layer any type of filter over another. We spoke of this in a mini blog on its release, but this time we are going to do some examples to illustrate the power of the new Excerpts tab!
Have you been in a project and wondered what themes seem to come up with people mentioning feelings of anger but didn’t code for it? Have no fear, you can still analyze it!
First, we gather up terms that indicate anger…be exhaustive, but efficient. For example, anger will cover anything that has anger in it (anger, angered, but not angry).
NOTE: Some search terms are common in parts of other words. For example, the term “mad” would not just include mad, maddened, madder, etc but also Madison which is clearly not a word related to anger. In those cases, you can add a space before and after the term to only include the single “mad”
Next, now that we have our terms, we can get right into creating the filter:
This will narrow your active excerpts to those that include the term anger in them, however you will still need to add excerpts that have the other terms in them. For each of those, follow these steps next:
Finally, the promised analysis! Any charts you open will have excerpts with those search terms in them. As such, if you go to the Packed Code Cloud chart for example, you see what codes were most prominent in the context of anger. For a more exact view of each code’s application counts, the code application chart is the place to be. To see how themes interacted with each other, check out the Code Co-Occurrence chart! If a particular combination of codes trends with anger, you will see it there.
Now, we know your time is valuable and as such we will do one more illustrative example using our existing filter as a base. Let’s say you want to further refine your search down to male native participants and specifically what they said in their second interview.
First, we need to talk about how our filters work with each other in Dedoose. Conceptually, the order in which you apply the different layers of a complex filter does not matter. This is due to each step of the filter’s set up being a paring down of an existing data set. If a data point belongs to the final filter, no step in the filter’s setup will remove the point from view.
In practice however, setting up a filter for more than one search term involves adding data to a filter, since it is effectively impossible to remove all excerpts that don’t include a word instead of just adding those that have that word in them. In the end this means we must have something of an order to our operations to be certain the filter is being applied the way we intend.
As such, we always suggest the following order:
With that massive aside out of the way, here are the steps:
Are there any other filters you would like to define or are interested in? Do you create filters from the Analyze tab? Our user guide has plenty of information on how filters work in other areas of Dedoose, check it out here.
We understand that these filters can be difficult to set up at times and we’re always eager to help should you run into an issue, don’t be afraid to contact us here at support.
Please check out our events page, the webinars are a fantastic way to learn about the various features of the Dedoose application. https://www.dedoose.com/Resources/Webinars
We’d love to hear from you! We are always looking for ways to improve, if you have a question, suggestion or comment please forward them to Dedoose Support. Our users are our most powerful resource, everything we do is about making sure the application is the best it can be for YOU!