MASD Recap

A company surveys customers with 20 questions about their shopping preferences. They use factor analysis to reduce the data. What is the main purpose of factor analysis in this context?

• To identify distinct customer segments based on preferences

• To predict future shopping behavior

• To group correlated variables into underlying dimensions

• To calculate the frequency of shopping preferences

A company wants to identify high-value prospects for a new marketing campaign. Which method would be inappropriate for this purpose?

• Cluster analysis

• RFM Analysis

• Logistic Regression

• Linear Regression

A streaming service uses collaborative filtering to recommend movies. Which of the following best describes collaborative filtering?

• It uses demographic data to recommend movies

• It finds patterns in user-item interaction data to recommend items

• It identifies product associations based on past transactions

• It creates segments to target similar users

A company wants to prioritize high-value customers for loyalty rewards. Which method is the most appropriate, and why?

• Cluster Analysis, because it segments customers based on similarity

• Logistic Regression, because it predicts customer loyalty

• RFM Analysis, because it scores customers based on recency, frequency, and monetary value

• Survival Analysis, because it measures customer lifetime

A marketing team segments customers using K-means clustering. What is the primary limitation of this method?

• It cannot handle missing values

• It requires the number of clusters to be predefined

• It cannot work with continuous variables

• It always produces overlapping clusters

A KM curve shows a steep drop at month 6, followed by a plateau until month 12. What does this indicate?

• Most subjects experience the event between month 6 and month 12

• The event occurs frequently around month 6 but not between months 6 and 12

• The risk of the event increases steadily over time

• The event occurs uniformly throughout the observation period

A telecom company wants to identify groups of customers with similar behavior and predict which customers are likely to churn. Which combination of methods should they use?

• Cluster Analysis for grouping and Logistic Regression for prediction

• Logistic Regression for both grouping and prediction

• Cluster Analysis for prediction and Logistic Regression for grouping

• Factor Analysis for grouping and Cluster Analysis for prediction

In a KM analysis comparing survival times for two treatments, the survival curves cross at month 8. What does this suggest?

• One treatment is always better than the other

• The effectiveness of treatments changes over time

• There is no difference between the treatments

• Both treatments fail at the same rate after month 8

A company analyzes the relationship between advertising spend and sales using linear regression. The model has an Adjusted R-squared of 0.85. What does this indicate?

• 85% of the variation in sales is explained by the model.

• • The advertising spend has an 85% impact on sales

• 85% of sales are predicted accurately.

• The model is overfitted.

In a regression model, the coefficient for Education Level is positive when included alone but becomes negative when combined with Income. What is the most likely explanation?

• Income and Education Level are highly correlated

• The dependent variable is poorly defined

• There is measurement error in Education Level

• The sample size is too small