MaxDiff reporting
When a survey containing a MaxDiff question is live, the reporting element "MaxDiff Summary" will be added by default to the Smart Report and can also be added to the Blank Reports.
Complete the steps in QuickView features in the ReportBuilder to view a report for a survey that contains a MaxDiff question. In the example below, the report is for a survey named "Credit Card Survey" that asked survey takers which features are most or least important to them when acquiring a new credit card. This question is listed at the top of the report.
Editing report elements
You can edit this report to include relative importance metrics, total counts of most and least important selections across all survey takers, as well as confidence intervals. To select the data elements that appear in the report, select Edit report in the top right corner of the report.
Select the report, which opens a panel containing Text, Settings, Data, and Style tabs (see Basic report element properties for details about these tabs). Select the Settings tab. Then select the items you want to see in the report. See below for descriptions of available options.
- Rank
- Based on the utility score; utility scores are sorted and used to calculate ranking.
- Utility score
- Calculated using the Empirical Bayesian estimation, this score is recalculated each time you change a specific filter in the report. Utility score is calculated using distribution percentage, least important count, most important count, and times the attribute was displayed to all survey takers.
- Relative importance
- Importance of specific attributes in relation to other attributes. Relative importance predicts the likelihood that a specific attribute will be chosen as most or least important on a scale of 1 to 100.
- Distribution %
- Percentage of the instances where a specific attribute was selected as most or least important.
- Standard error
- Denotes the accuracy of the relative importance value as compared to its likely true value in the overall population. The higher this value, the more likely that the relative importance score is an inaccurate representation of its true value in the population.
- Times displayed
- Number of times a specific attribute was presented to survey takers.
- Least important
- Total number of times a specific attribute was chosen as least important by all survey takers.
- Least important proportion
- Number of times the attribute was chosen as Least Important, divided by the number of times the attribute was shown.
- Most important
- Total number of times a specific attribute was chose as most important by all survey takers.
- Most important proportion
- Number of times the attribute was chosen as Most Important, divided by the number of times the attribute was shown.
- Confidence internal
- The 95% confidence interval is a range of values that one can be 95% confident contains the true mea of the population. Both the Upper Confidence and Lower Confidence levels are displayed; these give the boundaries within which the true value of the estimate is likely to exist.
Filtering report data
To filter data that appears in the report, select the filter icon located in the top right corner of the report. Make your selections and click Apply.
When you apply filters, data is recalculated and MaxDiff modeling and analysis is run based on the filters selected. You can choose to export the report as an image or as an Excel download after each filter is applied. To do this, select the options menu (three vertical dots) in the top right corner of the report and select either Save as or Save as spreadsheet.
