Background

In a bustling corporate office, Sarah, a finance manager at Apex Financial Solutions, was tasked with optimizing the collections process. With a portfolio of clients from various industries, the company’s success relied heavily on its ability to recover overdue payments. Sarah suspected that the effectiveness of the phone scripts used by collections associates varied, but no one had analyzed which script drove the best results.

Sarah’s goal was to determine the most profitable script among four options: Script A, Script B, Script C, and Script D. However, the monetary collections data she received were highly skewed due to outliers like large payments from a few clients, making traditional parametric analysis unsuitable. She decided to use Mood’s Median Test, a non-parametric statistical method, to compare the scripts’ performance.

Parametric hypothesis tests assume that the data follows a specific distribution, usually normal, and rely on parameters like the mean and standard deviation. They are powerful when these assumptions hold true, providing precise and efficient estimates. However, when data are non-normal, contain outliers, or violate other assumptions, parametric tests can yield misleading results. Non-parametric tests, like Mood’s Median Test, do not require specific distributional assumptions and are based on the ranks or medians of the data. They are more robust and flexible, making them suitable for analyzing skewed or ordinal data. Choosing between the two depends on the data’s characteristics and the validity of underlying assumptions.

What They Did

Data Collection and Preparation

Sarah gathered three months of monetary collections data. Each collection call was tagged with the script used and the dollar amount collected. To anonymize the data, Sarah labeled the scripts as A, B, C, and D and prepared a dataset. Below is the median dollar collected for each of the scripts used.

The dataset showed clear evidence of non-normality: while many calls resulted in small or zero-dollar collections, a few produced exceptionally high payments.

Outcomes

Analyzing the Data Using Mood’s Median Test

To perform the Mood’s Median Test, Sarah used Minitab _R statistical software. She organized the data into four groups based on the scripts and inputted the collection amounts into Minitab _R . Here’s the output she received:

Mood’s Median Test: Dollar Collected versus Script

Data Summary

Group    N      Median

Script A 120      150

Script B 110      200

Script C 115       75

Script D 125    250

Mood’s Median Test

Chi-Square = 45.67

DF = 3

P-Value < 0.001

Interpretation:

Since the p-value is less than 0.05, there is significant evidence to suggest that the median collections differ among the scripts.

The test revealed a statistically significant difference in the median collections across the four scripts. To identify the best-performing script, Sarah reviewed the group medians:

• Script D had the highest median at $250.
• Script B followed with $200.
• Script A had a median of $150.
• Script C had the lowest median of $75.

Further Analysis and Decision-Making

Sarah didn’t stop at the medians. She conducted a post-hoc pairwise comparison to confirm that Script D was statistically better than the others. The results reinforced her initial finding: Script D was the most profitable.

Implementing Changes

Sarah presented her findings to the collections department. She proposed replacing all other scripts with Script D and training associates to deliver it effectively. To test her hypothesis, Sarah launched a three-week pilot program where all associates exclusively used Script D.

Evaluating the Results

After the pilot, Sarah compared the weighted dollar collections using all four scripts ($150) before the change and after implementing only using Script D. Here’s the estimated annual improvement data:

The median collection amount increased by $100, and total collections rose by $800,000—a 53% improvement.

Final Thoughts

The transition to Script D proved transformative for Apex Financial Solutions. With higher collections and greater efficiency, the company increased profitability while maintaining positive client relationships. Sarah’s analytical approach, leveraging Mood’s Median Test, showcased the power of non-parametric statistics in real-world decision-making.

Her efforts also inspired other departments to adopt data-driven strategies, cementing Sarah’s reputation as a leader in innovation. This story demonstrates how a finance manager like Sarah used a non-parametric statistical approach to identify the most profitable collections script, leading to significant business improvements. By choosing the right statistical tools and leveraging data effectively, organizations can uncover valuable insights even from non-normal data distributions.

The image featured at the top of this post is ©Ground Picture/Shutterstock.com.