Pool Filters
You can filter on absolutely everything - Metadata, Questions from Screeners, Incentive Spending, Payment History, Email History, Scheduling History, Tags, and more. Please note there are separate articles for Combo Filters and Grouping Filters. If you have one filter, it will look like this:
If you have multiple filters, you'll see the filters in this dropdown:
1. Getting started
Click a + Add Filter at the top of Pool, and search for the name of a question in any screener or CSV upload. From this long list of potential filters, you can also find metadata filters like "Completed Last Interview" or "Last Paid Date." All you have to do is select a filter, which will take you to the next step below.
2. Filter options
Each option will allow you to filter your participants using absolute and relative boolean logic.
What does that mean? You can create a filter on data that is fixed or is comparable and join filters together with words like and, or and not.
For example, searching for responses before or after a fixed date and time—or, searching for responses less than 5 days ago/from today.
3. Grouping Filters
Creating strings of filters allows you to get more specific and narrow down your searches - please go here for all the details on grouping filters.
4. Filtering Questions with Multiple Answers
Some researchers call these "multi-select" questions, but think about any checkbox question or even scheduling question that might have multiple answers.
Things get a little more complicated when you start filtering around multiple data points / multi-select data. This happens when you have more than one answers for the same question. Think about it like a question about favorite fruit that contains multiple answers: apples, oranges, and pears. If you set a fruit filter to show apples, it will also show people that like apples and oranges.
This only affects certain kinds of filters. For example, using Payment History, if you are trying to "filter by someone who has a paid date before 2019." Yes, you'll see results from participants that have a paid date before 2019, but not exclusively. They could also have a paid date after 2019 and show, if they've been paid more than once and one happens to fall after 2019—in the same way as you can pick apples and oranges and have two answers for the same question.