Ma and Sun (2020)

Which of the following is highlighted as a key strength of machine learning methods compared to traditional statistical and econometric models typically used in marketing research?

Proven capacity to capture individual consumer-level heterogeneity and dynamics.Ability to readily handle unstructured data such as text, image, audio, and video.Excelling in the area of causal identification and interpretation.Providing a transparent and easily interpretable model structure.

It is from the discussion on the relative advantages of ML methods over econometric models, particularly concerning data type.

Section 2.3.1, paragraph 1: "The first key strength of machine learning methods is that they can readily handle unstructured data such as text, image, audio, and video, and can process data with complex structures such as large scale network or tracking data."

• False
• True
• False
• False

Which of the following is highlighted as a key strength of machine learning methods compared to traditional statistical and econometric models typically used in marketing research?

Ability to readily handle unstructured data such as text, image, audio, and video.Providing a transparent and easily interpretable model structure.Excelling in the area of causal identification and interpretation.Proven capacity to capture individual consumer-level heterogeneity and dynamics.

It is from the discussion on the relative advantages of ML methods over econometric models, particularly concerning data type.

Section 2.3.1, paragraph 1: "The first key strength of machine learning methods is that they can readily handle unstructured data such as text, image, audio, and video, and can process data with complex structures such as large scale network or tracking data."

• True
• False
• False
• False

Which of the following is highlighted as a key strength of machine learning methods compared to traditional statistical and econometric models typically used in marketing research?

Providing a transparent and easily interpretable model structure.Ability to readily handle unstructured data such as text, image, audio, and video.Excelling in the area of causal identification and interpretation.Proven capacity to capture individual consumer-level heterogeneity and dynamics.

It is from the discussion on the relative advantages of ML methods over econometric models, particularly concerning data type.

Section 2.3.1, paragraph 1: "The first key strength of machine learning methods is that they can readily handle unstructured data such as text, image, audio, and video, and can process data with complex structures such as large scale network or tracking data."

• False
• True
• False
• False

According to the provided framework (Fig. 1), which machine learning task is generally characterized by the training dataset containing only the input variables (X) without the corresponding output variables (Y)?

Unsupervised LearningActive LearningReinforcement LearningSupervised Learning

Review the definitions of the two major traditional task categories: supervised vs. unsupervised learning.

Section 2.1.2, paragraph 1: "In unsupervised learning tasks, the training dataset contains only the input variables, while the output variables are either undefined or unknown. The typical goal is to find hidden patterns in or extract information from the data." (Also, see Fig. 1 under "Tasks").

• True
• False
• False
• False

According to the provided framework (Fig. 1), which machine learning task is generally characterized by the training dataset containing only the input variables (X) without the corresponding output variables (Y)?

Supervised LearningUnsupervised LearningActive LearningReinforcement Learning

Review the definitions of the two major traditional task categories: supervised vs. unsupervised learning.

Section 2.1.2, paragraph 1: "In unsupervised learning tasks, the training dataset contains only the input variables, while the output variables are either undefined or unknown. The typical goal is to find hidden patterns in or extract information from the data." (Also, see Fig. 1 under "Tasks").

• False
• True
• False
• False

According to the provided framework (Fig. 1), which machine learning task is generally characterized by the training dataset containing only the input variables (X) without the corresponding output variables (Y)?

Supervised LearningUnsupervised LearningReinforcement LearningActive Learning

Review the definitions of the two major traditional task categories: supervised vs. unsupervised learning.

Section 2.1.2, paragraph 1: "In unsupervised learning tasks, the training dataset contains only the input variables, while the output variables are either undefined or unknown. The typical goal is to find hidden patterns in or extract information from the data." (Also, see Fig. 1 under "Tasks").

• False
• True
• False
• False

In the introductory example of a customer's e-book reader purchase journey, which action is performed by an AI agent using a reinforcement learning-type approach

The consumer reading product reviews on a third-party website.The consumer searching for e-readers at Google.The consumer receiving a coupon in email to finalize the order.A chatbot answering the consumer's questions via live-chat.

Reinforcement learning is suited for automated, context-dependent decisions that interact with the environment. Chatbots exemplify split-second, human-like conversations. While several actions in the scenario are AI-guided (like search ranking, ad delivery via bidding machines), the text explicitly mentions chat-bots answering questions in the context of automated systems making context-dependent decisions. Additionally, the text mentions chat-bots engage human-like conversations with customers to maintain relationship and loyalty.

• False
• False
• False
• True

In the introductory example of a customer's e-book reader purchase journey, which action is performed by an AI agent using a reinforcement learning-type approach

A chatbot answering the consumer's questions via live-chat.The consumer searching for e-readers at Google.The consumer receiving a coupon in email to finalize the order.The consumer reading product reviews on a third-party website.

Reinforcement learning is suited for automated, context-dependent decisions that interact with the environment. Chatbots exemplify split-second, human-like conversations. While several actions in the scenario are AI-guided (like search ranking, ad delivery via bidding machines), the text explicitly mentions chat-bots answering questions in the context of automated systems making context-dependent decisions. Additionally, the text mentions chat-bots engage human-like conversations with customers to maintain relationship and loyalty.

• True
• False
• False
• False

In the introductory example of a customer's e-book reader purchase journey, which action is performed by an AI agent using a reinforcement learning-type approach

A chatbot answering the consumer's questions via live-chat.The consumer receiving a coupon in email to finalize the order.The consumer reading product reviews on a third-party website.The consumer searching for e-readers at Google.

Reinforcement learning is suited for automated, context-dependent decisions that interact with the environment. Chatbots exemplify split-second, human-like conversations. While several actions in the scenario are AI-guided (like search ranking, ad delivery via bidding machines), the text explicitly mentions chat-bots answering questions in the context of automated systems making context-dependent decisions. Additionally, the text mentions chat-bots engage human-like conversations with customers to maintain relationship and loyalty.

• True
• False
• False
• False

A digital marketing firm is attempting to use millions of unstructured social media comments to segment its customer base for highly personalized ad campaigns. Which advantage of machine learning is most critical for this task?

Focus on causal identification, ensuring accurate ROI measurement.Scalability and ability to process large-scale, unstructured data.Lack of model transparency, allowing for proprietary algorithms.Reliance on traditional statistical models like linear regression

The scenario involves "millions of unstructured social media comments," emphasizing both volume and data type.

• False
• True
• False
• False

A digital marketing firm is attempting to use millions of unstructured social media comments to segment its customer base for highly personalized ad campaigns. Which advantage of machine learning is most critical for this task?

Focus on causal identification, ensuring accurate ROI measurement.Reliance on traditional statistical models like linear regressionLack of model transparency, allowing for proprietary algorithms.Scalability and ability to process large-scale, unstructured data.

The scenario involves "millions of unstructured social media comments," emphasizing both volume and data type.

• False
• False
• False
• True

A digital marketing firm is attempting to use millions of unstructured social media comments to segment its customer base for highly personalized ad campaigns. Which advantage of machine learning is most critical for this task?

Scalability and ability to process large-scale, unstructured data.Lack of model transparency, allowing for proprietary algorithms.Reliance on traditional statistical models like linear regressionFocus on causal identification, ensuring accurate ROI measurement.

The scenario involves "millions of unstructured social media comments," emphasizing both volume and data type.

• True
• False
• False
• False

Which of the following is identified as a major limitation of many machine learning methods from the perspective of academic marketing research?

Relationships uncovered are often correlational instead of causal.Requirement of unrealistically small, structured datasets for training.Inability to generate accurate out-of-sample predictions.Over-reliance on simple linear functional forms.

Section 2.3.2: "Second, relationships uncovered using machine learning methods are often correlational rather than causal... With a predictive focus, little attention has been paid to endogeneity concerns when developing machine learning methods."

• True
• False
• False
• False

Which of the following is identified as a major limitation of many machine learning methods from the perspective of academic marketing research?

Requirement of unrealistically small, structured datasets for training.Relationships uncovered are often correlational instead of causal.Inability to generate accurate out-of-sample predictions.Over-reliance on simple linear functional forms.

Section 2.3.2: "Second, relationships uncovered using machine learning methods are often correlational rather than causal... With a predictive focus, little attention has been paid to endogeneity concerns when developing machine learning methods."

• False
• True
• False
• False

Which of the following is identified as a major limitation of many machine learning methods from the perspective of academic marketing research?

Over-reliance on simple linear functional forms.Relationships uncovered are often correlational instead of causal.Requirement of unrealistically small, structured datasets for training.Inability to generate accurate out-of-sample predictions.

Section 2.3.2: "Second, relationships uncovered using machine learning methods are often correlational rather than causal... With a predictive focus, little attention has been paid to endogeneity concerns when developing machine learning methods."

• False
• True
• False
• False

In the Al-driven marketing landscape (Fig. 2), the trends of "Interactive & media-rich," "Personalization," "Real-time automation," and "Customer-journey focus" directly drive which of the following?

AI & Machine Learning methodology.All of the above.Marketing Actions (e.g., Search, Recommendation).Marketing Mix and Attribution.

Check Figure 2.

• False
• True
• False
• False

In the Al-driven marketing landscape (Fig. 2), the trends of "Interactive & media-rich," "Personalization," "Real-time automation," and "Customer-journey focus" directly drive which of the following?

AI & Machine Learning methodology.All of the above.Marketing Mix and Attribution.Marketing Actions (e.g., Search, Recommendation).

Check Figure 2.

• False
• True
• False
• False

In the Al-driven marketing landscape (Fig. 2), the trends of "Interactive & media-rich," "Personalization," "Real-time automation," and "Customer-journey focus" directly drive which of the following?

Marketing Actions (e.g., Search, Recommendation).Marketing Mix and Attribution.AI & Machine Learning methodology.All of the above.

Check Figure 2.

• False
• False
• False
• True

According to the conceptual framework for leveraging machine learning (Fig. 4), which two usages of machine learning methods are considered underexplored areas that require the highest combination of "Method Transparency" and "Marketing Theory Connection"?

Causal interpretation and Prescriptive analysis.Prediction and Feature extraction.Prediction and Estimation/optimizationDescriptive interpretation and Estimation/optimization.

Check Figure 4.

• True
• False
• False
• False

According to the conceptual framework for leveraging machine learning (Fig. 4), which two usages of machine learning methods are considered underexplored areas that require the highest combination of "Method Transparency" and "Marketing Theory Connection"?

Descriptive interpretation and Estimation/optimization.Prediction and Feature extraction.Prediction and Estimation/optimizationCausal interpretation and Prescriptive analysis.

Check Figure 4.

• False
• False
• False
• True

According to the conceptual framework for leveraging machine learning (Fig. 4), which two usages of machine learning methods are considered underexplored areas that require the highest combination of "Method Transparency" and "Marketing Theory Connection"?

Prediction and Estimation/optimizationCausal interpretation and Prescriptive analysis.Prediction and Feature extraction.Descriptive interpretation and Estimation/optimization.

Check Figure 4.

• False
• True
• False
• False

In the conceptual framework, "Injecting human insights" is a key consideration for which major aspect of empirical marketing research?

UsageDataTheoryIssue

Human insights and domain knowledge are crucial for connecting the abstract nature of ML methods to the generalizable results required by marketing theory.

• False
• False
• True
• False

In the conceptual framework, "Injecting human insights" is a key consideration for which major aspect of empirical marketing research?

IssueTheoryDataUsage

Human insights and domain knowledge are crucial for connecting the abstract nature of ML methods to the generalizable results required by marketing theory.

• False
• True
• False
• False

In the conceptual framework, "Injecting human insights" is a key consideration for which major aspect of empirical marketing research?

UsageTheoryIssueData

Human insights and domain knowledge are crucial for connecting the abstract nature of ML methods to the generalizable results required by marketing theory.

• False
• True
• False
• False

The shift toward real-time optimization and automation in marketing is a key industry trend driven by AI. What is the primary reason why firms are compelled to remove human agents from the "critical path" of marketing decision-making?

Consumers dislike human interaction and strongly prefer automated chatbots for all service needs.The complexity of the marketing environment, combined with fine-grained segmentation and the need for immediate response, surpasses manual human capacity.Human agents prefer to focus only on creative design and strategy, not execution.Legal regulations now prohibit human agents from handling personalized pricing decisions.

There are sections that explain why automation became necessary for marketing operations.

• Section 3.1.3: "The complexity of marketing environment has long surpassed the threshold of human analysts' intuitive understanding and manual capacities ... Fine-grained segmentation and frequent customer interactions further make it necessary to remove human agents from the critical path ... When mobile tracking detects an inbound consumer... a window of only minutes exists to deliver a promotional offer.

• False

• True

• False

• False

The shift toward real-time optimization and automation in marketing is a key industry trend driven by AI. What is the primary reason why firms are compelled to remove human agents from the "critical path" of marketing decision-making?

Consumers dislike human interaction and strongly prefer automated chatbots for all service needs.Legal regulations now prohibit human agents from handling personalized pricing decisions.Human agents prefer to focus only on creative design and strategy, not execution.The complexity of the marketing environment, combined with fine-grained segmentation and the need for immediate response, surpasses manual human capacity.

There are sections that explain why automation became necessary for marketing operations.

• Section 3.1.3: "The complexity of marketing environment has long surpassed the threshold of human analysts' intuitive understanding and manual capacities ... Fine-grained segmentation and frequent customer interactions further make it necessary to remove human agents from the critical path ... When mobile tracking detects an inbound consumer... a window of only minutes exists to deliver a promotional offer.

• False

• False

• False

• True

The shift toward real-time optimization and automation in marketing is a key industry trend driven by AI. What is the primary reason why firms are compelled to remove human agents from the "critical path" of marketing decision-making?

The complexity of the marketing environment, combined with fine-grained segmentation and the need for immediate response, surpasses manual human capacity.Human agents prefer to focus only on creative design and strategy, not execution.Consumers dislike human interaction and strongly prefer automated chatbots for all service needs.Legal regulations now prohibit human agents from handling personalized pricing decisions.

There are sections that explain why automation became necessary for marketing operations.

• Section 3.1.3: "The complexity of marketing environment has long surpassed the threshold of human analysts' intuitive understanding and manual capacities ... Fine-grained segmentation and frequent customer interactions further make it necessary to remove human agents from the critical path ... When mobile tracking detects an inbound consumer... a window of only minutes exists to deliver a promotional offer.

• True

• False

• False

• False

As firms increasingly deploy AI agents (driven by machine learning) to execute marketing tasks like retargeting and recommendation, what is a theoretical implication that researchers using non-experimental data must account for?

The AI's success ensures that a linear model will always outperform a complex deep learning model when assessing the firm's data.The use of AI introduces endogeneity concerns, meaning advertising or recommendation data must be analyzed while taking into account the underlying automated targeting and response policies.The deployment of AI eliminates all data silos, making unobserved heterogeneity irrelevant for causal studies.AI adoption guarantees that all historical data collected before the AI's deployment are invalid and cannot be used for any future research.

The paper explicitly states, "With firms using Al to power their business decision makings, the practice of adopting Al tools may have theoretical implications to marketing research. For example, the use of machine learning methods by practitioners raises endogeneity concerns for studies that use non-experimental data. If advertising data are obtained from a firm that performs retargeting, researchers have to take that into account when using the data to analyze consumers' ad response and to develop targeting strategies."

• False
• True
• False
• False

As firms increasingly deploy AI agents (driven by machine learning) to execute marketing tasks like retargeting and recommendation, what is a theoretical implication that researchers using non-experimental data must account for?

AI adoption guarantees that all historical data collected before the AI's deployment are invalid and cannot be used for any future research.The deployment of AI eliminates all data silos, making unobserved heterogeneity irrelevant for causal studies.The use of AI introduces endogeneity concerns, meaning advertising or recommendation data must be analyzed while taking into account the underlying automated targeting and response policies.The AI's success ensures that a linear model will always outperform a complex deep learning model when assessing the firm's data.

The paper explicitly states, "With firms using Al to power their business decision makings, the practice of adopting Al tools may have theoretical implications to marketing research. For example, the use of machine learning methods by practitioners raises endogeneity concerns for studies that use non-experimental data. If advertising data are obtained from a firm that performs retargeting, researchers have to take that into account when using the data to analyze consumers' ad response and to develop targeting strategies."

• False
• False
• True
• False

As firms increasingly deploy AI agents (driven by machine learning) to execute marketing tasks like retargeting and recommendation, what is a theoretical implication that researchers using non-experimental data must account for?

AI adoption guarantees that all historical data collected before the AI's deployment are invalid and cannot be used for any future research.The AI's success ensures that a linear model will always outperform a complex deep learning model when assessing the firm's data.The deployment of AI eliminates all data silos, making unobserved heterogeneity irrelevant for causal studies.The use of AI introduces endogeneity concerns, meaning advertising or recommendation data must be analyzed while taking into account the underlying automated targeting and response policies.

The paper explicitly states, "With firms using Al to power their business decision makings, the practice of adopting Al tools may have theoretical implications to marketing research. For example, the use of machine learning methods by practitioners raises endogeneity concerns for studies that use non-experimental data. If advertising data are obtained from a firm that performs retargeting, researchers have to take that into account when using the data to analyze consumers' ad response and to develop targeting strategies."

• False
• False
• False
• True