Business Statistics and Quantitative Methods
Group Project
Professor: Era Nagpal
Group Members:
1. Parthik Bajaj – 10357963
2. Saurav Shrestha – 10338719
3. Ravneet Kaur –
Part A: Data Collection
1. Business Case: To analyze customer feedback and understand service speed
performance, 15 customers were surveyed.
2. Each customer rated the service speed on a scale of 1 to 10, where 1 indicates
extremely poor service speed and 10 indicates excellent service speed.
Part B: Descriptive Statistics
Descriptive Statistics Table:
Statistic Value
Mean 7.00
Median (Q2) 7.00
Mode 7
Range 5
Variance 1.80
Standard 1.34
Deviation
Q1 (1st Quartile) 6.00
Q2 (2nd Quartile) 7.00
Q3 (3rd Quartile) 8.00
Interpretation:
● Central Tendency: These three central tendency measures (mean, median, and
mode) being equal or very close suggests that the data is symmetrical, with no
major skew — a strong indicator of consistency in customer experience.
● The majority of customers (75%) rated the service between 6 and 9, which
shows strong customer satisfaction with very few low outliers (like 4 or 5).
● The standard deviation and variance are low, indicating that the ratings are
tightly grouped — a sign of consistent service performance.
● The symmetry between mean, median, and mode means there are no extreme
biases or outliers drastically affecting the data.
● How these data influence business decisions:
● Reinforce Consistency: Since customer ratings are consistently high (around
7), the business should maintain its current service speed strategy.
● Address the Bottom 25%: A small segment (Q1 = 6 or less) shows reduced
satisfaction. Management could explore reasons behind these lower ratings (e.g.,
peak-hour delays) and introduce targeted improvements.
● Benchmarking: With a solid average of 7, the business can use this as a
benchmark to measure future changes or service upgrades.
● Marketing & Promotion: High satisfaction ratings can be used in promotional
messaging to attract new customers or justify premium pricing.
Part C: Frequency Distribution and Data Visualization
1. Frequency Distribution table
Cumulative
Class Interval Frequency Class Midpoint Frequency
4 – 4.99 1 4.5 1
5 – 5.99 3 5.5 4
6 – 6.99 7 6.5 11
7 – 7.99 10 7.5 21
8 – 8.99 7 8.5 28
9 – 9.99 2 9.5 30
Histogram
Interpretation:
● The histogram shows that the most common service speed ratings fall in the
range 7–7.99, with 10 customers giving ratings in this interval.
● The second most frequent ranges are 6–6.99 and 8–8.99, each with 7 ratings,
showing that most customers gave moderately high ratings.
● Very few customers rated below 6 (especially only 1 rating in 4–4.99), which
indicates high overall satisfaction.
● The distribution appears slightly left-skewed due to a small number of low
ratings (4 or 5), but it's largely concentrated around 7.
● The trendline indicates a slight upward trend, although it’s visually minor and
possibly influenced by grouping.
Frequency Polygon
Interpretation:
● The frequency polygon shows a clear peak at Class 4 (7–7.99), confirming that
this is where customer satisfaction is the highest.
● The curve is symmetrical, indicating low variability and a consistent service
experience.
● A steep rise from Class 1 to Class 4, followed by a balanced decline, suggests
predictable and repeatable service speed performance.
● No extreme spikes or dips indicate no major anomalies or outliers in customer
perception.
Part D: Probability Analysis
1. Probability-Based Questions
Q1: What is the probability that a customer gives a satisfaction rating of 7 or higher?
Ans: From the frequency distribution:
Ratings of 7–7.99: 10 customers
Ratings of 8–8.99: 7 customers
Ratings of 9–9.99: 2 customers
Total = 10 + 7 + 2 = 19 customers
P(Rating ≥ 7) = 19 / 30 = 0.6333 or 63.33%
Q2: What is the probability that a randomly selected customer gave a rating below 6?
Ans: From the distribution:
Ratings of 4–4.99: 1 customer
Ratings of 5–5.99: 3 customers
Total = 1 + 3 = 4 customers
P(Rating < 6) = 4 / 30 = 0.1333 or 13.33%
Q3: What is the probability that a customer gave a rating between 6 and 8 (inclusive)?
Ans: From the distribution:
Ratings of 6–6.99: 7 customers
Ratings of 7–7.99: 10 customers
Ratings of 8–8.99: 7 customers
Total = 7 + 10 + 7 = 24 customers
P(6 ≤ Rating ≤ 8) = 24 / 30 = 0.8 or 80%
2. Conditional Probability
Q4: If a customer gave a rating of at least 6, what is the probability that they gave a
rating of 8 or higher?
Customers with rating ≥ 6: 7 (6–6.99) + 10 (7–7.99) + 7 (8–8.99) + 2 (9–
9.99) = 26 customers
Customers with rating ≥ 8: 7 (8–8.99) + 2 (9–9.99) = 9 customers
P(Rating ≥ 8 | Rating ≥ 6) = 9 / 26 ≈ 0.3462 or 34.62%
3. Interpretation of Probability Analysis
Majority Satisfaction: Over 63% of customers rated the service 7 or higher. This
reinforces a high level of satisfaction.
Low Dissatisfaction: Only 13% rated the service below 6, showing minimal
dissatisfaction.
Core Satisfaction Zone: 80% of ratings fall between 6 and 8, highlighting the business's
strength in delivering consistent service.
Upsell Potential: Among customers already rating the service 6 or higher, there's a 35%
chance they gave a rating of 8 or more — indicating a solid segment for loyalty
campaigns or upselling premium services.
Part E: Correlation and Regression
Adding another variable - Number of visits
Customer Rating Number of Visits
1 8 1
2 7 2
3 9 3
4 6 2
5 7 3
6 8 4
7 6 2
8 5 4
9 9 3
10 7 5
11 4 4
12 8 3
13 6 5
14 7 6
15 9 3
16 8 6
17 7 5
18 5 4
19 6 5
20 8 4
21 9 6
22 6 5
23 7 6
24 7 7
25 8 6
26 5 7
27 6 6
28 9 6
29 7 7
30 8 8
1. Correlation Analysis
Variables Selected:
X (Independent Variable): Number of Visits
Y (Dependent Variable): Rating
Correlation Coefficient (r):
r =0.699
Interpretation: The correlation is positive (0.699), indicating a moderate to strong
relationship. As the number of visits increases, the customer rating tends to increase.
Since, r is closer to 1 than to 0, there is a meaningful linear association between visits
and ratings.
2. Regression Analysis
Regression Equation:
Rating = 5.392 + 0.358 × (Number of Visits)
Rating=5.392+0.358×(Number of Visits)
Key Values:
Intercept (5.392): If a customer visits 0 times, the predicted rating is 5.392
Slope (0.358): For each additional visit, the rating increases by 0.358 points on average.
R-squared (0.488): ~48.8% of the variation in ratings is explained by the number of
visits.
Interpretation:
The model shows that more visits lead to higher ratings, but other factors also influence
ratings.
The slope is statistically significant, meaning visits have a real impact on ratings.
Prediction Example:
If a customer visits 5 times, their predicted rating is:
5.392 + 0.358 × 5 = 7.182 ( ≈7.2)
Part F: Conclusion and Recommendation
1. Conclusion
The statistical analysis of customer feedback and visit frequency reveals key insights
about service performance and customer satisfaction:
a) Central Tendency & Dispersion:
The mean, median, and mode ratings were all 7, indicating symmetrical data distribution
and consistent service quality.
Low standard deviation (1.34) confirms that ratings are clustered closely around the
mean, with minimal variability.
b) Probability Analysis:
63.3% of customers rated the service 7 or higher, while only 13.3% rated it below 6.
80% of ratings fell between 6 and 8, highlighting a strong core satisfaction zone.
c) Correlation & Regression:
Positive correlation (r = 0.699) between number of visits and ratings: Frequent visitors
tend to rate the service higher.
Regression model: Each additional visit increases ratings by 0.358 points, explaining
48.8% of rating variability.
Key Business Implications:
The business delivers reliable service speed, but there’s room to improve satisfaction
among infrequent visitors.
Loyalty programs could leverage the link between visits and higher ratings.
2. Recommendations
a) Enhance Service for Low-Rating Segments:
Investigate why 13.3% of customers rated service below 6 (e.g., peak-hour delays, staff
training gaps).
Implement targeted feedback surveys for these customers to identify pain points.
b) Leverage Customer Loyalty:
Since more visits correlate with higher ratings, introduce:
Reward programs (e.g., discounts for repeat visits).
Personalized offers to encourage frequent visits.
c) Improve Predictive Analytics:
Use the regression model (Rating = 5.392 + 0.358 × Visits) to:
Predict satisfaction levels for new customers.
Identify "at-risk" customers (e.g., those with declining visit frequency) for retention
campaigns.
d) Benchmark & Monitor:
Track ratings monthly against the current mean (7.0) to detect shifts in service quality.
Compare performance across branches or time periods to isolate best practices.