<aside> 💡 The Phoenix platform will make all the data you gather available in an editable dashboard, with any classifications you have added applied to it. Phoenix dashboards are built on Superset, a powerful, open-source data visualisation tool.

</aside>

Navigating your dashboard

When you click on “Go to dashboard” in the Explore tab of the Phoenix platform, you will be taken to your dashboard, which will initially show a few graphs, tables and filters that we think may be useful as you start to look for insights in the data you have gathered (and maybe classified).

• Your dashboard is automatically connected to the standard table we described in the gathered data output section of this manual. All of the data for all of your gathers is in one consolidated table, which is connected to this dashboard.
• You have complete control over your dashboard, meaning that you can edit or delete all the elements that are there, and add new ones. This includes adding / editing the tabs, headings, filters, graphs and tables. Click on EDIT DASHBOARD in the top left corner to get started.
• If you add more data or add / edit a classifier, your dashboard will be automatically updated. If you can’t see the changes in your dashboard, then click on the three dots next to EDIT DASHBOARD and then “Refresh dashboard”.
• The dashboard is not public; it is password protected and connected to your Phoenix account. You can export individual charts as images or CSV files for use in reports and presentations.

Tips for editing your dashboard

There are many options available to edit the dashboard. Below, we share some ideas to start off with and a couple of tips specific to the way data is structured in Phoenix.

Filters

You can apply filters to your charts or tables to focus on specific data segments. We have created a few filters, which you can use to see what they do to your charts and tables. Every time you select a filter, it will apply to all of your charts and tables. To add or edit filters, click on “+ ADD/EDIT FILTERS” at the top of the filters.

<aside> 💡

Filter example:

• My data has author classes “journalist” and “politician”, and post classes “violence” and “protest”
• When I apply the filter “journalist”, I can look at the graph “Post by class” and I will see how many posts that are about “violence” or “protest” are being posted by authors classified as “journalist”. </aside>

Charts

We have created a few charts that we think you will find useful, and you can create new charts by clicking on EDIT DASHBOARD and then on + CREATE NEW CHART. You will first need to select a dataset; the only dataset that appears in the dropdown menu is the correct one for your project. You can then select the type of chart you want to make. There are many options available to you; here are some ideas:

• Bar Charts: Useful for comparing quantities across different categories. For example, visualize the volume of posts by different classes.
• Pie Charts: Show proportions and percentages within a whole. For example, visualize the volume of posts by different accounts. Careful! Pie charts can be misleading in some cases — e.g. you cannot use pie charts to show the volume of posts by class, because a post may have more than one class.
• Line Charts: Ideal for showing trends over time. For example, track the frequency of specific classes of posts over a selected period.
• Heat Maps: Provide a visual representation of data density or intensity in a cross-reference. For example, create a heatmap of post class against comment class, visualising the density of certain comments in certain posts.

<aside> <img src="/icons/thought-dialogue_blue.svg" alt="/icons/thought-dialogue_blue.svg" width="40px" />

We’ve noticed that the structure of the standard data output table in Phoenix can make it hard to make certain charts. The main issue to keep in mind is that if you either (a) collect comments or (b) create classes, then you will have more than one row per unique post (and comment). This means that:

• Whenever you are making a chart that counts posts or comments, you should use the COUNT_DISTINCT function on either post_id or comment_id
• Whenever you are making a chart that adds up attributes related to posts or comments (e.g. number of likes), you should use the SUM(DISTINCT(xxx)) function. Although it is possible that you would miss some attributes (e.g. if posts have exactly the same number of likes), this will give you the right answer most of the time
• The AVERAGE function won’t work to compare attributes; we recommend using the MAX function instead for comparisons </aside>

Layout Elements