Overview
For Projective Mapping, panelists are asked to organize multiple samples in a two-dimensional space according to perceived similarities and differences. Panelists may be asked to tag the samples, for example, to associate a series of samples with a common attribute or descriptor. Panelists can be presented with sample labels and/or sample images.
In this workflow, we will focus on setting up a Projective Mapping test with the following characteristics:
- White background 650x433 pixels in size.
- Sample blinding codes displayed to panelists.
- Panelists can map the sample in any order they wish.
Panelists can create their own tags in addition to the ones you create.
It is important that panelists are instructed not to use 'special' characters, such as commas, slashes, etc., when creating their own tags. The 'special' characters can cause the analysis not to run. These characters then need to be removed manually in the raw results, which can be a very time consuming process prone to human error.
Test Setup
- From the Dashboard screen, select the Create a new test.
- Choose the test type Standard test type and click Use.
- Update the Test name.
- Locate and highlight the folder or sub folder you wish the test to reside in.
- Click Create test.
- In the Overview screen, update any of the following components as necessary: Test name , Time zone , Language , Website , Tags, and Description.
- In the Samples & design area of your test, add samples, add a design and update blinding codes if necessary.
- In the Build area, update the Welcome screen and Thank you screen text as necessary.
Add a Projective Mapping question.
Update the instructions to panelists. Click into the text editor to type instructions for your panelists.
Click into the question and select Question options. Update the following as necessary:
General. Update the general options for teh question, such as question and display name.
Display. Update the display options as needed, such as the Show question and Present question options.
Show blinding code = Yes
Choice display = Choice label only
Enforce evaluation in sample set order = No
Tags.
Click the + Add tag to add the tags that your panelists will use to label the samples.
In the tag Text column, type the text that will display to panelists. To delete a tag, click it in the list and then click X.
Set Allow Respondent to Create Tags = Yes
Background.
Keep the Background image as White and the Mapping area dimensions: Rectangle (650x433).
Click Save once done.
- Click Panelists to add panelists into the test.
- Review your options in Logistics, Preview and Run your test.
- Once results are done being collected, Pause your test and analyze data.
Projective Mapping Analysis
Projective Mapping data does not have a built in analysis report within Compusense Cloud. The data needs to be exported and analyzed in SensoMineR.
The graphs generated in R include:
- Individual graphs for each panelist
- Groups Representation
- Individual Factor Map
- Correlation Circle
- Confidence ellipses for the napping configuration
Installing SensoMineR
Projective mapping data from Compusense Cloud can be exported in R format. SensoMineR can be used to analyze the data as it provides numerous graphical outputs. SensoMineR is completely free and can be downloaded by following the link below.
http://sensominer.free.fr/
Scroll down on the main page to Install SensoMineR and follow the instructions. See image below.
In the Results area, select Reports > Create report.
In 1. Select report type, select Projective mapping data export.
In
2. Select options, select
.r export. In 3. Select questions, your Projective mapping question will be selected by default.
In 4. Select export type, select .r or .r (unicode encoding) if your language require this option.
Click the down arrow to save the file.
Generating Graphs in R
Open the R on your computer.
Click File > Open script .
Browse to the location on your computer where you saved the *.r export from Compusense Cloud.
Highlight the file and click Open . The R Editor window will open up.
Load the data for all graphs, by following these steps in the R Editor :
- Right-click and select Select All .
Hold down the Shift key on your keyboard and press the up arrow key on the keyboard three times to deselect the last three lines of code. See image below.
- Right-click in the selected/highlighted area and select Run line or selection . It will seem like nothing happened, but this was a necessary step to perform in order to generate the graphs outlined below.
Graph Types
Individual Graphs for Each Panelist
The Individual Graphs use each panelist’s raw data to plot the placement of products on the mapping area.
In the R Editor , highlight the third line from the bottom.
Right-click in the selected area and select Run line or selection .
Each window contains up to 4 panelist plots. See example image. If you had more than four panelists/results in the test, you will get more than one graph window.
The graph windows will be hidden behind the first graph window. Click on it and drag to the left to see the next graph window.
To copy the graph(s) into Excel, right-click on the graph and select Copy as bitmap . Open an Excel file and paste in it.
Groups Representation Graphs
Keep the R Editor window open, but the graphs previously generated should be closed before following the steps below:
In the R Editor window, highlight the second line from the bottom.
Right-click in the selected area and select Run line or selection . The Groups Representation graph will generate. See example image.
To copy the graph into Excel, right-click on the graph and select Copy as bitmap . Open an Excel file and paste in it.
Do not close the other graphs just yet. They are described further down on this page. You may minimize them to find the remaining Groups Representation graphs, as seen in the example image below.
Individual Factor Map
These are example Individual Factor Maps.
Correlation Circle Graphs
There are 3 correlation circles in the Correlation Circle graph:
Correlation of the assessor dimensions relative to the consensus space.
Correlation for the profiling terms relative to the consensus space.
- The superimposition of assessor dimensions and profiling terms in the consensus space.
- Repeat the copying process until you copy all the graphs.
- Close the graphs, but leave the R Editor open.
Confidence Ellipses for the Napping Configuration Graph
This graph gives a sense of how stable the configuration is. Confidence intervals are based on data from virtual panels.
The configuration of products is based on results from the true panel. Each virtual panels’ mean results are Procrustes-transformed onto the real configuration. The coordinates of all virtual panels are used to obtain confidence ellipses which investigate the uncertainty of the real product configuration.
This results in a 95% confidence ellipses for the product configuration. Big ellipses with lots of overlap indicates low stability of the configuration. Small ellipses with little overlap indicates high stability of the configuration.
In the R Editor window, highlight the bottom line.
Right-click in the selected area and select Run line or selection . Confidence ellipses for the napping configuration graph (single page) will generate. See example image.
- To copy the graph into Excel, right-click on the graph and select Copy as bitmap . Open an Excel file and paste in it.
Other Projective Mapping Data Exports
In addition to exporting into SensoMineR, there are two other Projective Mapping exports available. These files export in *.csv format, which is useful for analysis in XLSTAT, for example.
- In the Results area, select Reports > Create report.
- In 1. Select report type, select Projective mapping data export.
- In 2. Select options, select: .
- Projective Mapping Coordinates
- Projective Mapping Tags
- In 3. Select questions, your Projective mapping question will be selected by default.
- In 4. Select export type, select .r or .r (unicode encoding) if your language require this option.
- Click Create my report.
- Click the down arrow to save the file.
Related Articles
Question Type: Tetrad
Overview Tetrad test is a discrimination test type that helps you determine if there is a difference between two products by asking panelists to group 4 samples into 2 groups of 2. Unlike most other question types, the Tetrad question is not ...
Question Type: Choose
Overview In a Choose question, panelists are presented with a group of choices or concepts and asked to pick one choice, all choices that apply, or a specific number of choices. There are three types of Choose questions available in Compusense20: ...
Question Type: Triangle
Overview In Triangle questions panelists are presented with a total of 3 samples, two of one product and one of another. They are then asked to taste the samples in front of them from left to right. After they have tasted all three samples they are ...
Question Type: Category
Overview Category questions present a quantitative scale with defined values and descriptions to allow panelists to make a selection based on their perception of a sample or concept. In Compusense20 we have made it easier for you to add a ...
Question Type: Sort
Overview In a Sort question, Panelists are presented with a group of samples or concepts and asked to organize them into bins. There are three types of Sort questions available: Free sort: The Free Sort question allows panelists to sort products into ...