Back to Previous

Analyzing Measured Response and Sensory Evaluation Data: Knowing If Your New Product or Service is Really Different

ENDED

Training by Technopoly Inc

Inquire Now

On-Site / Training

PHP 11,200.00

Inquire Enroll Now

RESPONSE RATE

Details
Outline
About the Provider

Enroll Now

Inquire

Details

Product testing is part and parcel of any company’s new product development process. In fact, no new product or service gets commercialized unless the test is done and the new product has been shown to be favorably different than its competitor or perhaps significantly preferred or accepted by the test’s chosen set of respondents. As critical as it is though, product testing response data is either not fully or correctly taken advantage off. This could happen in various ways but the two most common are – (1) running the analysis primarily through descriptive summary statistics (e,g, averages) or graphical depiction of results which can highly limit the insight that can be gained or worse lead off to a wrong one; (2) applying the wrong statistical analysis approach (e.g. using the popular T-test to rank or rated data) which can give an erroneous conclusion.

Truth is, the proper statistical analysis approach to a product test or response data will differ depending on what measure is being used and studied as well as the number of treatments involved. Treatments could refer to alternative formulations, brands, designs, etc. Not only that, but also of important consideration is how the test was carried out such as independent or dependent (paired) testing of treatments, monadic or sequential monadic application, complete or incomplete treatment set respondent assessment. Taking into account these considerations, the statistical techniques to be covered will deal with the cases as follows:

Testing and analyzing difference between treatments (independent or paired) with objective measured data
Testing and analyzing difference between treatments (independent or paired) with subjective measure or sensory (rating/ranking) data with sensory fatigue considerations (incomplete treatment assessments)
Testing and analyzing preference between treatments (e.g. triangle or duo-trio tests) or acceptance of a treatment

This learning session will primarily be delivered through lecture considering the technical nature of the topics involved. It will follow a cycle of lecture then computational workshop for each statistical analysis tool and if time allows- a management of learning in the sense of how each participant foresees the potential application of the tools learned to their work or projects.

HOW YOU WILL BENEFIT

Develop the ability to identify and use the proper statistical inferential analysis tool contingent on the product testing design and project/study intent.
Correctly analyze, interpret and draw insights from product and design testing results.
Structure product and sensory evaluation test aligned to the development project/study intent and not be constrained by a limited set of statistical analysis tool to use.

To reserve for this learning session or should you have any questions, please feel free to contact us via email through [email protected] with your name and contact details.

Outline

Day 1

AM Session

1.0 Brief Primer on Statistical Analysis

1.1 Scales of Data Measurement

1.2 Statistical Inference

1.3 Parametric vs Non-Parametric Statistical Test

1.4 Statistical Summaries

1.5 Graphical Data Display

2.0 Analyzing and Testing for Measured (Objective) Data

2.1 Comparing or Differentiating Two Brands or Formulations

Independent and Paired T-Tests
Comprehensive Look at Data: Use Both Average and Standard Deviation

PM Session

F-Test for Equality of Data Variance
Alternative for Small Sample Size: Wilcoxon Rank-Sum and Signed Rank Test

2.2 Comparing or Differentiating Two or More Brands of Formulations

Review of One-Way ANOVA with and without Blocks
Test of Multiple Comparison Tools: Duncan’s Multiple Range Test and Tukey’s Test

Day 2

AM Session

3.0 Analyzing and Testing for Rated or Ranked (Subjective) Data

3.1 Testing for Two Brands or Formulations: Review of Rank-Sum and Signed Rank Test

3.2 Testing for Multiple Brands or Formulations:

Friedman Test for Respondent Complete Treatments Application
Durbin Test for Respondent Incomplete Treatments Application
Kruskal-Wallis for Monadic Testing

PM Session

Page Test for Ordered Multiple Alternatives
Nemenyi’s Joint Rank and Dunn’s Tests

3.3 Estimating Single Proportion and Difference of Two Proportion Values (Support for Top X Boxes Approach)

4.0 Analyzing and Testing for Preference (Subjective) Data

4.1 Testing for Two Brands or Formulations: Mc Nemar’s Test

4.2 Testing for Multiple Brands or Formulations: Cochran’s Q Test

Speaker/s

FACILITATOR/MENTOR

Bryan Gobaco

Bryan has trained companies on a number of product research, development and testing topics and tools. These companies are in the field of consumer goods and personal care, agri-business, specialty chemicals, plastics and packaging, food ingredients to name some. He has also done work on market research and new product survey for personal care and finance company. He is a resident trainer of the Philippine Trade Training Center and a lecturer of the Department of Industrial Engineering at the Gokongwei College of Engineering of De La Salle University Manila where he also earned his bachelor in science and masters in science degree in Industrial Engineering.

Reviews

Be the first to write a review about this course.

Write a Review

We are a management advisory and consulting group focused on process and product development and improvement methodology, business analytics and decision support modeling services.

We believe in supporting process excellence initiatives of our client-partners through the creation of a clear strategic and operational mindset that emphasizes on customer and business value.

View Complete Profile

Chemical Engineering Food Processing Marketing & Advertising