XLSTAT: MaxDiff Analysis in Excel

 

Background

Dataset For Example

The (attached) data represent a simple simulated case of MaxDiff analysis based on tourism marketing. 
A hotel brand wishes to open a new hotel and would like to understand the importance of attributes of a vacation destination in order to define its offer.

The results are divided into different sheets:

1. Attributes: The attributes selected to describe the products.

2. Designs for MaxDiff: The generated design, and the choices given by 5 respondents.

3. MaxDiff: The results of MaxDiff analysis by hierarchical Bayes

 

Setting Up

Step 1 of MaxDiff analysis involves selecting key attributes—here, eight were chosen with tourism experts (e.g., beaches, culture, cuisine).

Step 2 uses XLSTAT to generate attribute sets for comparison. By selecting Designs for MaxDiff analysis under the CJT menu, XLSTAT creates optimized sets using a design of experiments approach to ensure balanced and efficient comparisons.

In the dialog box, enter the analysis name, select the attribute names, set the number of comparisons (e.g., 8), and define the number of profiles per comparison (e.g., 4).

These settings balance data quality and respondent attention.

In the Outputs tab, select Combination / Respondent format to display combinations in rows and respondents in columns. Leave Print individual sheets unchecked, as it's not needed here—though useful for collecting responses in detailed studies.

After clicking OK, XLSTAT generates the MaxDiff sheet with a choice table.

This table is to be filled after interviews, with each individual having two columns: one for their most important (best) and one for their least important (worst) attribute per set, using values from 1 to 4.

Step 3 Fill in the MaxDiff Analysis Table

MaxDiff tables can be filled manually after external interviews or automatically via individual sheets with linked responses. Using individual sheets simplifies data entry, which is helpful given the complexity of completing the main table directly.

For combination C1, individual 1 selects Entertainment as the most important and Wildlife close by as the least important attribute among the four shown.

 

Interpretating the Results

After collecting responses from 5 individuals in the MaxDiff sheet, you can run the analysis by either clicking Run the analysis or by selecting the MaxDiff feature from the CJT menu in XLSTAT.

You can then select the data.

To auto-load the Responses and Combinations datasets, click the magic stick icon. In the pop-up window, select any cell in the original MaxDiff design sheet (e.g., H13), then click OK. This works if the design was generated in XLSTAT and hasn't been manually altered.

Manually select your data as follows:

• Responses field: Select the 10 columns containing respondent answers.

• Choices table: Select the 4 columns that represent the attribute details.

In the Options tab, select Hierarchical Bayes and keep all other settings default. For a quicker analysis, choose only Counts in the Outputs tab.

Once you click the OK button, the computations are performed and the results are displayed. 

The MaxDiff scores are displayed as rescaled coefficients summing up to 100. We see that Individual 1 thinks nightlife and luxury hotel offer are very important. Shopping around and wildlife are also important but to a lesser extent.

Then the descriptive statistics are shown. If we check the Means column, we see that wildlife around is the most important attributes on average.

 

We can visualize average MaxDiff scores on the chart below.

After this preliminary study, we would advise the hotel chain to focus on nightlife in a luxury hotel with cultural activities.