Session 3 - Statistical and Descriptive Analytics

Step 1: Start with the Right Business Questions

Before calculating numbers, it is essential to approach your dataset with a business mindset. Simply running statistics without purpose does not lead to useful insights.

For the Amazon Sales Dataset, here are some relevant business questions you might consider:

• Which product categories are the most popular among customers?

• What are the minimum, maximum, and average prices for the products?

• Are the discounts offered having an impact on customer satisfaction?

• Which products are receiving poor customer ratings, and why?

• Which products are performing well, and what can we learn from them?

By thinking through these questions, you can focus your statistical analysis on the areas that truly matter for business decision-making.

Step 2: Prepare Your Stats Sheet

To keep your analysis organized, create a new sheet within your Google Sheets file. Name this new sheet Stats. This is where you will calculate and document your key statistical metrics.

Start by creating column headers in the Stats sheet:

Step 3: Calculate Minimum, Maximum, and Average Ratings

Let’s begin by analyzing the customer ratings from your Amazon dataset. The ratings are located in Column G of your main data sheet.

To calculate the Minimum Rating:

1. In the Stats sheet, under “Value” next to “Minimum Rating,” enter:
   =MIN(Amazon!G:G)

2. Press Enter.
   This will display the lowest customer rating across all products.

To calculate the Maximum Rating:

1. In the next row under “Value,” enter:
   =MAX(Amazon!G:G)

2. Press Enter.
   This shows the highest rating given to any product.

To calculate the Average Rating:

1. In the following row, enter:
   =AVERAGE(Amazon!G:G)

2. Press Enter.
   This gives you the overall average rating across all products.

For example, you may find that the minimum rating is 2, the maximum is 5, and the average is approximately 4.1. These numbers already give you a sense of how your products are performing from the customer’s perspective.

Step 4: Measure Variability with Standard Deviation

Average values alone do not tell the full story. You also want to know how much individual ratings differ from the average. This is called standard deviation, and it helps you understand how consistent or spread out your ratings are.

To calculate Standard Deviation:

1. In the Stats sheet, in the next available row, type:
   =STDEV(Amazon!G:G)

2. Press Enter.

A low standard deviation means most of your product ratings are close to the average. A high standard deviation suggests that there are wide differences — some products performing much better or worse than others.

For example, if your standard deviation is small, and the average rating is 4.1, you can be confident that most products are rated close to that number.

Step 5: Identify Low-Rated Products

It is important to know which products received low ratings so that action can be taken to improve them.

To do this:

1. Go to your Amazon Sales Dataset sheet.

2. Select the entire dataset.

3. Click on Data → Sort Range → Advanced Range Sorting Options.

4. Ensure “Data has header row” is checked.

5. Sort by Rating (Column G) from A to Z.

This will bring the products with the lowest ratings to the top of your sheet.

You can now see, for example, how many products received a rating of 2 or 2.5, and whether these low-rated products are isolated cases or part of a broader pattern.

To count these low-rated products:

• Select the cells with a rating of 2 to 2.9.

• Look at the count displayed at the bottom right of Google Sheets.

You can repeat this for products with ratings of 3 to explore products that may need improvement but are not performing as poorly as those with the lowest ratings.

Step 6: Explore Relationships with Correlation

Beyond individual statistics, you can explore how different factors in your dataset relate to one another. For example, you might ask:

Does offering a higher discount lead to better customer ratings?

To check for a relationship between discounts and ratings, use the CORREL function.

To calculate correlation between Discount and Rating:

1. In the Stats sheet, enter:
   =CORREL(Amazon!F:F, Amazon!G:G)

2. Press Enter.

The result will be a number between -1 and 1:

• A value close to 1 indicates a strong positive relationship — as one increases, so does the other.

• A value close to -1 suggests a strong negative relationship — as one increases, the other decreases.

• A value near 0 indicates little or no relationship between the two.

In our case, you get a correlation of -0.15, this suggests that there is no significant relationship between discounts and customer ratings in your dataset. This means that simply offering discounts does not appear to influence how customers rate the products.

Step 7: Key Takeaways

With just a few simple formulas, you have performed a basic yet powerful statistical analysis:

✔ You calculated minimum, maximum, and average ratings to assess product performance.
✔ You used standard deviation to understand the variability in customer satisfaction.
✔ You identified poorly rated products that may require attention.
✔ You explored potential relationships between discounts and customer ratings.

Statistical analysis transforms raw data into actionable business insights. Even with basic tools like Google Sheets, you can answer meaningful questions and guide decision-making.

What’s Next?

In the upcoming session, we will dive deeper into the dataset using Exploratory Data Analysis (EDA). You will learn how to visually explore your data to uncover hidden patterns, outliers, and trends.