1 of 15

Enabling Informed Consumer Choice

with Simplified AHP�Dr. Bill Adams and Dr. Jay Stryker

2 of 15

The Problem

  • Too many options
  • Marketers work hard!
  • How can consumers navigate?

3 of 15

The Solution

  • Break into smaller questions
  • Simple answers
  • Combine into overall

4 of 15

The Theory

  1. AHP Structured�
  2. Likert Pairwise for Consumer�
  3. Combined With Expert Ratings�
  4. Giving overall alignment

5 of 15

The Implementation

  • Google forms questionnaire�
  • Google sheets for data storage�
  • Javascript: Consumer results�
  • Jupyter+Python: Deep dive

6 of 15

The Implementation (in a picture):

7 of 15

Theory: AHP Structure

  1. Break decision into a hierarchy (nodes)�
  2. Original choices are at the bottom (alternatives)�
  3. Rate and Weight

8 of 15

Theory: AHP Pairwise

Numerically Rank?

​

  • Cost�
  • Style�
  • Functionality

​

Too hard, so we:

Pairwise

Gives Ranks

Cost: 0.6

Style: 0.3

Funct: 0.1

9 of 15

Theory: Simplified AHP Pairwise Using Likert

And we convert the words to a pairwise vote

10 of 15

Theory: AHP Expert Alternative Ratings

With many alternatives:

  1. Too many pairwise comparisons�
  2. But perhaps experts can evaluate

In our model

Experts can evaluate a Phone on:

  • Cost
  • Style
  • Photography
  • Gaming
  • Storage

​

11 of 15

AHP Theory

Recap for our model

  1. Consumers Pairwise Criteria
    1. Cost
    2. Style
    3. Functionality
      1. Photography
      2. Gaming
      3. Storage�
  2. Experts Rate Phones on Criteria�
  3. sAHP to Synthesize Results

12 of 15

Research Work

  1. Adjusting the anpjs library to work with Google Forms�
  2. Adjusting pyanp for Google Forms�
  3. Designing the AHP Model�
  4. Creating Expert Ratings�
  5. Creating results website�
  6. Group Analysis

13 of 15

Research Work

Dr. Bill Adams

  • Adjusting the anpjs library to work with Google Forms�
  • Adjusting pyanp for Google Forms�
  • Designing the AHP Model�
  • Creating Expert Ratings�
  • Creating results website�
  • Group Analysis

14 of 15

Research Work

Dr. Jay Stryker

  • Adjusting the anpjs library to work with Google Forms�
  • Adjusting pyanp for Google Forms�
  • Designing the AHP Model�
  • Creating Expert Ratings�
  • Creating results website�
  • Creating Analysis.ipynb

15 of 15

Source code available:

  1. Consumer Choice AHP: https://github.com/wjladams/consumerAHP�
  2. Javascript ANP Library: https://github.com/dlens/anpjs �
  3. Python ANP Library: https://github.com/wjladams/pyanp