MaxDiff

Example 1

Summary: Perform MaxDiff analysis using a Firth bias-adjusted logistic regression model to analyze multiple-choice data with profile effects.

Code:

// Open data table
dt = Open("$Sample_Data/Design Experiment/Candy Survey.jmp");
// MaxDiff
MaxDiff(
    One Table( 1 ),
    Response Subject ID( :Subject ),
    Profile ID( :Choice ),
    Profile Grouping(
        :Subject, :Choice Set
    ),
    Profile Effects( :Candy ),
    "Firth Bias-adjusted Estimates"n( 1 )
);

Example 2

Summary: Perform One Table MaxDiff analysis with Firth Bias-Adjusted Estimates on data from Potato Chip Combined.jmp.

Code:

// Open data table
dt = Open("$Sample_Data/Potato Chip Combined.jmp");
// One Table MaxDiff
MaxDiff(
    One Table( 1 ),
    Subject ID( :Respondent ),
    Choice Set ID( :Choice Set ID ),
    Profile ID( :Response ),
    Profile Grouping( :Survey ID ),
    Profile Effects( :Profile ID ),
    "Firth Bias-Adjusted Estimates"n( 1 ),
    Response Value Indicates Best( 1 ),
    Response Value Indicates Worst( -1 )
);

Example 3

Summary: Perform a MaxDiff analysis for flavor preference using the Potato Chip Responses, Profiles, and Subjects data tables .

Code:

// Open data table
dt = Open("$Sample_Data/Potato Chip Responses.jmp");
// MaxDiff for Flavor
Open(
    "$Sample_Data/Potato Chip Subjects.jmp"
);
Open(
    "$Sample_Data/Potato Chip Profiles.jmp"
);
MaxDiff(
    Response Data Table(
        Data Table(
            "Potato Chip Responses"
        )
    ),
    Profile DataTable(
        Potato Chip Profiles
    ),
    Subject DataTable(
        Data Table(
            "Potato Chip Subjects"
        )
    ),
    Response Subject ID( :Respondent ),
    Response Profile ID Choices(
        :Choice 1, :Choice 2, :Choice 3
    ),
    Profile ID( :Profile ID ),
    Profile Effects( :Flavor ),
    Subject Subject ID( :Respondent ),
    Subject Effects(
        :Citizenship, :Gender
    ),
    "Firth Bias-adjusted Estimates"n( 1 ),
    Response Best Option( :Best Profile ),
    Response Worst Option(
        :Worst Profile
    )
);

Example 4

Summary: Perform MaxDiff analysis with no subject effects using the MaxDiff platform.

Code:

// Open data table
dt = Open("$Sample_Data/Potato Chip Responses.jmp");
// MaxDiff with No Subject Effects
Open(
    "$Sample_Data/Potato Chip Subjects.jmp"
);
Open(
    "$Sample_Data/Potato Chip Profiles.jmp"
);
MaxDiff(
    Response Data Table(
        Data Table(
            "Potato Chip Responses"
        )
    ),
    Profile DataTable(
        Potato Chip Profiles
    ),
    Subject DataTable(
        Data Table(
            "Potato Chip Subjects"
        )
    ),
    Response Subject ID( :Respondent ),
    Response Profile ID Choices(
        :Choice 1, :Choice 2, :Choice 3
    ),
    Profile ID( :Profile ID ),
    Profile Effects( :Flavor ),
    Subject Subject ID( :Respondent ),
    "Firth Bias-adjusted Estimates"n( 1 ),
    Response Best Option( :Best Profile ),
    Response Worst Option(
        :Worst Profile
    )
);

Example 5

Summary: Analyze customer preference data for potato chips using Max Diff with a focus on product of value.

Code:

// Open data table
dt = Open("$Sample_Data/Potato Chip Responses.jmp");
// Max Diff for Product Of
Open(
    "$Sample_Data/Potato Chip Subjects.jmp"
);
Open(
    "$Sample_Data/Potato Chip Profiles.jmp"
);
MaxDiff(
    Response Data Table(
        Data Table(
            "Potato Chip Responses"
        )
    ),
    Profile DataTable(
        Potato Chip Profiles
    ),
    Subject DataTable(
        Data Table(
            "Potato Chip Subjects"
        )
    ),
    Response Subject ID( :Respondent ),
    Response Profile ID Choices(
        :Choice 1, :Choice 2, :Choice 3
    ),
    Profile ID( :Profile ID ),
    Profile Effects( :Product Of ),
    Subject Subject ID( :Respondent ),
    Subject Effects(
        :Citizenship, :Gender
    ),
    "Firth Bias-adjusted Estimates"n( 1 ),
    Response Best Option( :Best Profile ),
    Response Worst Option(
        :Worst Profile
    )
);

Example 6

Summary: Perform a MaxDiff analysis using Hierarchical Bayes with additional effects from Profile and Subject data tables, and apply Firth Bias-Adjusted Estimates.

Code:

// Open data table
dt = Open("$Sample_Data/Potato Chip Responses.jmp");
// MaxDiff with Hierarchical Bayes
Open(
    "$Sample_Data/Potato Chip Subjects.jmp"
);
Open(
    "$Sample_Data/Potato Chip Responses.jmp"
);
Open(
    "$Sample_Data/Potato Chip Profiles.jmp"
);
MaxDiff(
    Response Data Table(
        Data Table(
            "Potato Chip Responses"
        )
    ),
    Profile DataTable(
        Data Table(
            "Potato Chip Profiles"
        )
    ),
    Subject DataTable(
        Data Table(
            "Potato Chip Subjects"
        )
    ),
    Response Subject ID( :Respondent ),
    Response Profile ID Choices(
        :Choice 1, :Choice 2, :Choice 3
    ),
    Profile ID( :Profile ID ),
    Profile Effects( :Product Of ),
    Subject Subject ID( :Respondent ),
    Subject Effects(
        :Citizenship, :Gender
    ),
    Hierarchical Bayes( 1 ),
    Hierarchical Bayes( 1 ),
    "Firth Bias-Adjusted Estimates"n( 1 ),
    Response Best Option( :Best Profile ),
    Response Worst Option(
        :Worst Profile
    )
);