This guide is designed to walk you through how to configure reporting in Watershed to answer key questions about how your organization uses Degreed. If you haven’t yet connected Degreed to Watershed, you can learn how to do that here.

How many people are logging in to Degreed?

Measure(s)

In order to know how many people are logging in to Degreed, we need to create a measure that filters out all data that isn't explicitly a user logging in to Degreed. Within Watershed's measure editor, the new measure configuration should resemble the screenshot below, focusing on the filter, value, and aggregation.

You can also configure this measure using the advanced configuration version of the Measure Editor. To do so, copy the configuration below into the advanced view:

"name": "Login Count",
 "aggregation": {
  "type": "COUNT"
 },
 "valueProducer": {
  "type": "STATEMENT_PROPERTY",
  "statementProperty": "id",
  "displayId": false
 },
 "filter": {
  "verbIds": {
   "ids": [
    "https://api.degreed.com/xAPI/verbs/loggedin"
   ]
  }
 }

Report

Because we want to know how many people are logging in to Degreed regularly over a period of time, we want to choose a visualization that represents a time series well. For our needs, the Line Report will work perfectly. From Watershed's Report Builder interface, select the Line report type and then choose the measure we've just created above. The "Organized" section allows us to choose to plot the number of logins over a given day, week, month, or year. Let's assume we want to see it for every day.

Once saved, we now have a report that quickly tells us at a glance for any given day how many logins have been made into Degreed:

What skills are most added in Degreed?

Measure(s)

If we want to know which skills are being added the most, we once again begin with creating a measure to filter out any data not related to skills in Degreed. Degreed captures this event using the verb "added skill". The full ID of the verb can also be used as "https://api.degreed.com/xAPI/verbs/addedskill". We see in the image below that we can create a measure very similar to our Login Count measure, except to change the verb filter used. 

What if we want to do more than just count the skills being added, but compare which skills are being added more over time? To do that, we can create two more measures that filter on the same verb and add additional date filters. Doing this requires using advanced configuration. Below is a screenshot highlighting where the date filter is added in the advanced configuration section to only count the skills that have been added within the past week.

You can even copy and paste the below configuration directly into your advanced configuration:

{
 "name": "Last Week",
 "aggregation": {
  "type": "COUNT"
  },
 "valueProducer": {
  "type": "STATEMENT_PROPERTY",
  "statementProperty": "id",
  "displayId": false
  },
 "filter": {
  "verbIds": {
   "ids": [
    "https://api.degreed.com/xAPI/verbs/addedskill"
   ]
  },
  "dateFilter": {
   "dateType": "trailing",
   "trailingAmount": 1,
   "trailingType": "weeks",
   "olderThanAmount": 0,
   "olderThanType": "none",
   "fieldName": "timestamp"
  }
 }
}

Report 

Now we're ready to implement our measures into a report in Watershed. For this example we're using the heatmap report since we'll want to quickly and easily see trends across multiple columns of data. Once we have the heatmap report type selected in Watershed's report builder, we can select the new measures we've created. In this example, I added another measure looking at skills added over the previous month as well as the previous week. For this report, we want to know what skills are most added, not what people are adding the most skills, so we choose to organize "By Activity". But as you can see, that leads to some wonky skill IDs showing up in the preview.

Because we want to see the human readable names of the skills and not their IDs, we need to use identify where in the Degreed xAPI statement that information is located, and then organize our data by that field. Using Watershed's Data Search screen we can isolate an xAPI statement describing the moment someone added a skill:

We can see that our default sorting by Activity is pulling back the ID instead of the full name. To sort by the correct field, we can use Advanced Configuration to change the dimension of the report. Using what we learned from data search, we can set the exact statement property we want to organize our report by as shown in full below.

"dimensions": [
 {
  "type": "STATEMENT_PROPERTY",
  "statementProperty": "object.definition.description.und",
  "caseSensitive": true,
  "displayId": false
 }
]

Now that we've sorted out that issue, we have a beautiful heatmap showing us which skills are being added the most in Degreed, broken down by their popularity over the past week, month, and all-time.

What are the hottest Pathways in Degreed?

Measure(s)

To determine what Pathways are the most popular in Degreed, we need to once again start with creating measures that will filter on the most relevant data to help us answer that question. Degreed captures this event using the verb "followed". The full ID of the verb can also be used as "https://api.degreed.com/xAPI/verbs/followed". You can see in the image below the basic configuration for this measure. To ensure we only look at data concerning Pathways, we need to add an additional filter as shown below the Verb filter. The second filter is checking to be sure that the type of every object we look at is a pathway - https://api.degreed.com/xAPI/activities/pathway.

We can now create a report similar to our Login Count report above. However, because we've decided we want to know which Pathways are trending, the best way to do so is to add additional date parameters, which will allow us to compare Total Follows against those over the past week or month. First, we copy our existing measure to a new measure renames "Pathway Follows In The Last Week". Keeping the same filters in place, we will add one more by copying the below dateFilter into the measure configuration.

"dateFilter": {
 "dateType": "trailing",
 "trailingAmount": 1,
 "trailingType": "weeks",
 "olderThanAmount": 0,
 "olderThanType": "none",
 "fieldName": "timestamp"
 }

 Report(s)

We can now configure a leaderboard to demonstrate which are the hottest Pathways in Degreed. To do so we'll select the Leaderboard report type from the Report Builder, and select the two measures we've configured above.

You'll notice in this report that we've added a third measure called "% Increase Over Previous Week". This measure is designed to really highlight for us which Pathways are becoming very popular based on the increasing rate at which they're being followed week over week. This measure was added to the report using a formula measure. In order to calculate the week over week rate of change, we have to get a little creative and use a bit of math.

The formula used to calculate this change by week is as follows:

(# Follows Last Week) / (# Follows Last 2 Weeks-#Follows Last Week)*100

 In order to use that formula we'll create two new measures. The formula measure itself as described above, and the measure used in the formula for the number of pathway follows in the last two weeks. To create the latter of the two, we can simply copy our Pathway Follows In The Last Week measure and update the name and filter to count follows for the previous two weeks instead of one. Once we've added that to our report, the formula measure is added via advanced configuration with the following syntax, where m0 is Follows Last Week and m1 is Follows Over Two Weeks. Note that we also apply some rounding functions to the formula as well so we have a neat result on the report.

{
 "name": "% Increase Over Previous Week",
 "aggregation": {
  "type": "FORMULA",
  "formula": "floor(m0/(m1-m0)*100)*100/100"
 },
 "valueProducer": {
  "type": "STATEMENT_PROPERTY",
  "statementProperty": "id",
  "displayId": false
  }
 }

This is just an introduction to the possible reporting that can be done with data from Degreed. We would love to hear how you've used Watershed to answer compelling questions about your learning in Degreed!