Overview
The Median Value calculator finds the middle value of a selected numerical attribute across all cases or events in your process. Unlike the average, the median is not affected by extreme outliers, making it a more robust measure of central tendency when your data contains unusual values. The median represents the 50th percentile - half of all values are below it, and half are above it.
Common Uses
- Calculate median invoice amounts when outliers might skew the average
- Determine median processing times for realistic performance expectations
- Find median payment timeliness to understand typical payment behavior
- Measure median order values for more accurate forecasting
- Identify the middle point of attribute distributions when data is skewed
- Add median value metrics to dashboards for outlier-resistant analysis
Settings
Attribute Name: Select the numerical attribute for which you want to calculate the median value. This can be any case or event attribute containing numeric data.
Examples
Example 1: Median Payment Timeliness
Scenario: Your accounts payable process has a few invoices with extremely early payments (special discounts) and some very late payments (disputed invoices). You want to understand typical payment behavior without these extremes skewing your results.
Settings:
- Attribute Name: PaymentTimeliness
Output:
The output shows a single value, such as "-5 days", indicating that the median payment is made 5 days early (negative values indicate early payment). This differs from the average of "-16 days" because a few very early payments (30-60 days early) were pulling the average down.
Insights: The median reveals that most payments are actually only slightly early, while the average was heavily influenced by a small number of exceptionally early payments. This gives you a more realistic picture of typical payment behavior and helps you set appropriate payment scheduling policies.
Example 2: Median Invoice Processing Time
Scenario: Your invoice processing times range from 1 day to 45 days, with most invoices processed within a week but a few stuck in approval loops. You want to know the typical processing time without being misled by these outlier cases.
Settings:
- Attribute Name: Case Duration
Output:
The calculator displays "3.10:24:15" (3 days, 10 hours, 24 minutes, 15 seconds), showing that half of all invoices are processed faster than this, and half take longer. This is much more representative than the average of 7 days, which was inflated by the few cases stuck for 30-45 days.
Insights: This median value helps you set realistic SLAs and customer expectations. You now know that typical cases complete in about 3.5 days, and can investigate the outliers separately without them distorting your understanding of normal operations.
Example 3: Median Order Value in E-commerce
Scenario: Your online store has order values ranging from $10 to $5,000. A few very large corporate orders are making the average order value misleading for marketing analysis.
Settings:
- Attribute Name: Total Order Value
Output:
The calculator shows a median of $87, while the average is $245. This reveals that most customers spend around $87, but a few large orders are dramatically increasing the average.
Insights: Use the median value of $87 for more accurate marketing decisions, such as setting free shipping thresholds or designing promotional offers. The average of $245 was misleading because it was skewed by a small number of high-value orders.
Example 4: Median Time to First Response
Scenario: Your customer service process tracks time from case creation to first response. Most responses occur within 2 hours, but some cases received during off-hours wait until the next business day.
Settings:
- Attribute Name: Time To First Response
Output:
The median value is "0.01:45:00" (1 hour, 45 minutes), showing that half of all customers receive their first response within this timeframe. The average was 4.5 hours due to overnight cases.
Insights: The median provides a better performance metric for your team's responsiveness during business hours. You can now set SLAs based on typical performance and handle after-hours cases separately.
Output
The calculator displays a single numerical value representing the median (middle value) of the selected attribute across all applicable cases or events. When the dataset has an even number of values, the median is calculated as the average of the two middle values.
The median is particularly valuable when:
- Your data contains outliers or extreme values
- You need a robust measure that's resistant to unusual cases
- You want to understand typical performance rather than overall performance
- Your data distribution is skewed rather than normally distributed
This documentation is part of the mindzie Studio process mining platform.