Build: FCM - Guidelines For Implementation

Build: FCM - Guidelines For Implementation

Overview

As you are considering implementing the Compusense® FCM® (Feedback Calibration Method) into your training routine, you might wonder, "How do I implement the FCM method for my panel training initiative?" or "How do I determine what to set the targets and ranges to!?"

These are great questions, we are glad you asked. :)  Let's explore!


Guidelines For Implementing FCM

Below are some general guidelines for implementing FCM.

If you would like to schedule with Compusense a formal, detailed training on implementing FCM as your training tool, please contact our Support Team.

  1. Select and screen panelists. Decide whether you will use internal or external panelists, trained or untrained. Then screen them for their recognition and learning ability, for ranking ability, and for discrimination ability.

  2. Implement basic sensory training techniques. Follow “Best Practices” for consistent, actionable results. Blind code the samples. Standardize the following: testing environment, method of sample preparation, method of sample evaluation.

  3. Develop the lexicon and organize the attributes. Understand the range of attributes and intensities. Conduct attribute generation on all or a representative subset of the test products, etc.

  4. Classify attributes and develop reference standards. Consider ease of identifying and scaling attributes. Provide reference standards and confirm their suitability. Use CATA for attributes not detected by all panelists. Develop and document clear attribute definitions.

  5. Create the ballot and practice scaling. Introduce the ballot to panelists and discuss.

  6. Collect data in Compusense using FCM. Use fixed sample order for training purposes. Data collected can be used in future feedback.

  7. Evaluate panelist performance using FCM reporting. Use our built-in Hits and misses report, and Panelist summary report for your training needs.

  8. Refine the feedback training plan. After the product evaluations, refine targets and ranges and monitor performance.

To answer the question "How do I determine what to set the targets and ranges to!?", we need to look at two different approaches:
  1. Feedback on products with existing profiles: Use historical data or Gold Standard data.

  2. Feedback on products without existing profiles: Generate attributes, select key attributes to initially focus on, use spikes to simulate the attribute range.


Both of these approaches are described in more detail below.


Determine The Targets And Ranges From Existing Profiles

The general guideline for determining targets and ranges from existing profiles is as follows:
  1. Select the key attributes to provide feedback on.

  2. Demonstrate low, mid-range and high target values.

  3. Use historical confidence intervals to set the range.

  4. Present the project with feedback.

  5. Debrief.

Here is an example calculation for targets and ranges based on the Summary report generated in Compusense:


The data in this example was collected on a 100-point unstructured Line Scale. Different colours in the image indicate significant difference between samples.

From the image above we can see that the mean value for Product 1 is 74. The Confidence interval (Tukey's HSD calculation in this example) is 11. The formulas are as follows:
  1. Lower Range Value = Mean - Confidence Interval

  2. Upper Range Value = Mean + Confidence Interval
As you are putting all these values together, it may be beneficial to save them in the correct format right away so that they are ready for import into Compusense. The easiest way to determine the right format is by exporting it from Compusense:
  1. Go to the Build tab and click the Targets & ranges button.

  2. Click Export, save the file, open it and populate with the target and range values you are compiling.


  3. When done, save the file by keeping the .csv file format. Follow the steps for importing the values into Compusense.


Determine The Targets And Ranges From Preliminary Feedback

Since there is no existing profiling data to work from:
  1. Your panel needs to generate attributes.

  2. The panel leader should select key attributes to initially focus on. Use Compusense Summary report to assist you in the decision making.

  3. Use spikes to simulate the attribute range.

  4. The panel leader must present the panelists with products that demonstrate low, mid-range and high intensities of the key attributes.

  5. From this initial evaluation, generate the preliminary target and range values. Set a meaningful target, a practical range, practice and refine the feedback. Use Compusense Summary report and raw data export to assist you with setting targets and ranges.


You are now ready to set the test up! Select the workflow that meets your training needs:






    • Related Articles

    • Build: FCM - Panelist View

      Overview As you are considering implementing the Compusense® FCM® (Feedback Calibration Method) into your training routine, you are probably wondering what will the panelist experience be. You have come to the right place. Below you can see a table ...

    • Build: FCM For Category Attributes

      Overview On your journey through the Compusense® FCM® (Feedback Calibration Method) you have arrived to the point where you want to use Category scale to train your panelists. If you have not already, please go through the steps leading up to this ...

    • Build: Immediate Feedback

      Overview Compusense® FCM® (Feedback Calibration Method) is a proven computerized training method developed by Compusense Inc. to train descriptive sensory panelists with immediate feedback to their responses. Panelists, panel leaders, and businesses ...

    • Build: FCM For Choose 1 Questions

      Overview On your journey through the Compusense® FCM® (Feedback Calibration Method) you have arrived to the point where you want to use Choose 1 questions to train your panelists. If you have not done it already, please go through the steps leading ...

    • Build: FCM For Line Scale Attributes

      Overview On your journey through the Compusense® FCM® (Feedback Calibration Method) you have arrived to the point where you want to use Line scale to train your panelists. If you have not done it already, please go through the steps leading up to ...