Skip to content

Latest commit

 

History

9 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 

Repository files navigation

⭑✮🛍️⊹ Understanding Consumer Behaviour Toward GenAI Shopping Assistants

Survey Analytics using SPSS, CHAID Decision Trees & Neural Networks

Why do some people enjoy using AI while shopping, trust its suggestions and want to continue using it, while others do not?

Is it because the assistant responds well? Feels human? Is reliable? Makes shopping more enjoyable?

This research project explores exactly that.

Conducted as part of my research work at NIT Tiruchirappalli, the study analyzes survey responses from 366 GenAI users to understand what influences their experience and intention to use AI for shopping.

The project uses survey analytics, reliability testing, descriptive statistics, CHAID Decision Trees and Multilayer Perceptron (MLP) to turn consumer responses into meaningful behavioural insights.

✦ The Survey

The study used a 5 point Likert scale survey with responses from 366 people who had used Generative AI.

The analysis focused on four main consumer outcomes:

😊 Satisfaction → How happy and satisfied people are with the GenAI shopping assistant

💭 Behavioural Intention (BI) → Whether a person is willing or plans to use a GenAI shopping assistant

🛍️ Patronage Intention → Whether the GenAI assistant makes people more willing to shop with or buy from the platform

🔄 Continuance Intention (CI) → Whether a person who has already used it intends to keep using it in the future

Simply put: BI = “Will I use it?” while CI = “Will I continue using it?”

👥 Who Responded?

A few key patterns stood out:

👨 Male respondents formed a larger share of the sample than female respondents

🎓 Students were the largest group, followed by private sector employees

📚 Most respondents were undergraduates, followed by postgraduates

🏙️ Urban residents formed the largest group

🤖 All participants included in the study had experience using Generative AI

🛍️ Shopping goods were the most commonly reported category, meaning products where people usually compare different options before buying

🔍 What Factors Were Studied?

The study tested 10 factors that could influence how people experience and respond to GenAI shopping assistants.

🛠️ Quality Factors

Tangibility → How clear and visually useful the AI shopping experience feels

Responsiveness → How quickly and helpfully the AI responds

Reliability → How dependable and consistent the AI feels

Assurance → How much confidence and trust the AI creates

Empathy → How well the AI seems to understand the user's needs

💬 Social Factors

Anthropomorphism → How human like the AI feels

Social Credence → The social value or approval connected with using the AI

TPR → How the user experiences the AI's social presence and interaction

🎮 Engagement Factors

Autotelic → How enjoyable using the AI feels by itself

Immersion → How involved or absorbed the user feels while interacting with it

🌳 CHAID Analysis

The main technique used was CHAID Decision Tree Analysis.

In simple words, CHAID helped answer one main question:

Which factors matter the most for each type of consumer behaviour?

Four decision trees were created to study the factors influencing each consumer outcome.

The analysis showed that different factors played important roles across Satisfaction, Continuance Intention, Patronage Intention and Behavioural Intention.

✨ What Did We Learn?

The analysis showed that there was no single factor that explained every consumer outcome.

Different aspects of the GenAI shopping experience contributed differently to satisfaction, intention to use, willingness to shop through the platform and intention to continue using the technology.

Simply put, people respond to GenAI shopping assistants based on a combination of how well they work, how they interact with users and how engaging the experience feels.

🧪 Reliability & Descriptive Analysis

Before building the models, the survey was checked to make sure the responses were reliable and meaningful.

The constructs showed acceptable to excellent reliability, supporting their use in the further analysis.

The descriptive analysis also provided an overall understanding of how respondents viewed the different aspects of GenAI shopping assistants.

🧠 MLP Neural Network

A Multilayer Perceptron (MLP) neural network was also used as an additional predictive method.

Think of it as a second way of checking the patterns in the data.

CHAID helped show clear decision paths and consumer groups, while MLP helped study more complex relationships between the factors.

✦ Project Workflow

366 Survey Responses
        ↓
Demographic Analysis
        ↓
Descriptive Statistics
        ↓
Reliability Testing
        ↓
CHAID Decision Trees
        ↓
Consumer Behaviour Insights
        ↓
MLP Neural Network Analysis

🛠️ Tools & Techniques

IBM SPSS Statistics • Survey Analytics • Likert Scale Analysis • Descriptive Statistics • Cronbach's Alpha • CHAID Decision Trees • Multilayer Perceptron (MLP)

✦ Why This Project Matters

GenAI shopping assistants are becoming part of online shopping, but simply adding AI does not mean people will enjoy, trust or continue using it.

This project uses real consumer survey data to understand what actually influences people's behaviour and shows how hundreds of survey responses can be turned into clear, useful insights through data analysis.

📌 Research Note: This project is part of ongoing research. The respondent-level dataset and detailed SPSS outputs are not publicly shared.

⊹ Author

Manogna

Research Project • NIT Tiruchirappalli

⭑✮ Turning consumer responses into meaningful behavioural insights ₊˚⊹

About

Survey analytics project using responses from 366 GenAI users to understand what makes people enjoy, trust, shop with, and continue using Gen AI shopping assistants. Used SPSS, CHAID decision trees, reliability analysis, and MLP neural networks.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors