When performing an analysis on your Audit data points, the Qualaris analysis system has a few options you can use to dig deeper into your data and view the output in different ways. The Start and End date to specify which dates to include observations from, then filters can be applied to further narrow the data by specifying which data to include. A group-by can be specified that defines how the data is then divided up. The chart type defines the specific visualization type to use, and the data can then be ordered by name or value.
Sophisticated data analysis can be achieved by using combinations of these building blocks.
The best way to understand these concepts is to open the Explore tool which gives you direct access to all of the different options. You'll be able to explore your data in many ways by clicking through these options. We like to say you can "ask your data questions" and then get an immediate answer.
Here are specifics for each of the different analytics options.
A measure is a specific type of calculation. Currently, there are 3 supported measures.
Defect-free Compliance tells us how many process audits had at least 1 defect.
For a more detailed explanation, check out this article.
Adherence Compliance is defined as the percent of individual compliance items marked as compliant out of all questions answered.
For a more detailed explanation, check out this article (https://help.qualaris.com/en/articles/4836009-understanding-adherence-compliance).
The Count is the total number of audits.
Group-by and Filter options
Specifying a group-by will cause the results to be split into chunks where each chunk is a unique value from the group.
What is the Compliance of each individual unit?
How many data points are collected on each day of the week?
Which units haven't been observed enough yet this month?
For a more detailed example, if we start out with a default Explore session from Jan 1 to today and look at the Count Measure we would effectively be asking, "How many data points were submitted so far this year?".
If we specify "Unit" as the group-by, we will now see the same data split into chunks where each chunk is the count for each unit. This is asking the question, "How many data points per unit were submitted so far this year?"
When combined with filters, this can be a very powerful way to drill down into your data looking for trends and insights.
Filters allow you to specify what data to include in your analysis, and which data should be ignored.
For example, let's say you wanted to see the overall Compliance for the 2nd floor. You could use a filter to select data from only the 2N, 2E, 2W units.
This works the same way if you want to isolate a single unit's performance, or if you want to filter by a specific shift, etc.
Pro tip: you can string multiple filters and group-bys together to isolate very specific results.
In this example, we might want to group-by Unit again to break out the individual unit performance of the second floor in a single graph.
Shows data as a single number value or as a set of bars if a group-by is set. Used when you want to see aggregate data from the entire Start and Stop time range.
Used to show data over time. Requires selecting a time granularity (year, month, week, or day). The data from the Start and Stop range will be grouped together into the selected granularity and then plotted as a line.
You can show multiple lines by specifying a group-by.
The table visualization type uses a table format to show the same data as the metric, bar, or line. Filtering and specifying a group-by function identically to the other chart types. Setting can be set to year, month, week, or day to see the same data as a line, and setting it to None will yield the same data as a metric or bar.
Used to visualize data's proportion.
Note: Pie is only available for the Count measure
A Pareto chart is typically used to show the biggest contributors of defects to a process. A bar chart will show the frequency of occurrence for each defect in descending order, and a line chart will show the cumulative percentage of defects. This format quickly shows the biggest contributor of issues which would be the highest impact to improve.
For our analytics system, the most common way to configure a Pareto chart is to show the checklist items where the answer was the non-compliant answer (usually a "no") as the defect bars. The way to configure this is to filter for non compliant answers ("Compliance Field State" is "Not Compliant") and then group-by "Compliance Field". See the image for what this looks like.
Note: Pareto is only available for the Count measure
When you have specified a group-by, you can then select the way the group of items is ordered. The options are:
Name of group item, ascending or descending (for example, order alphabetically by Unit name)
Value of Measure, ascending or descending (for example, lowest count to highest count)