For Preference testing (where there is no right or wrong answer), please do not use the 2-AFC test template. Instead, use the Standard test type and add a Preference question in it.

For a back to back 2-AFC discrimination test, or a 2-AFC discrimination test with additional questions, please contact our Support Team to discuss your testing objectives and the best solution.

Question Compatibility

1. Category
   
   
2. Choose
   
   
3. Line Scale
   
   
4. Rank
   
   
5. Sort
   
   
6. Comment
   
   
7. Constant Sum
   
   
8. LMS
   
   
9. Numeric
   
   
10. Photo Upload
    
    

Text/Time Delay screens are compatible with all test types and question types .

1. On the Dashboard , in the Folders list on the left, select the folder where you wish to save your new 2-AFC test.
   
   
   If necessary, use the menu from the vertical three dots to create new sub-folders or to edit existing ones.
   
   
   
2. Click Create a new test .
   
   
3. Next to the 2-AFC test type , click Use .
   
   
   
   
   
4. Type the test name and specify the location for it (in case the folder selected in the first step was not the right one).
   
   
5. Click Create Test . The test will open up.
   

The Test ID displayed in the top right-hand corner is an important unique identifier useful for test searches. The Test ID can also be found in the Dashboard .

Please have the Test ID handy when you reach out to our Support Team for assistance.

1. Overview
   

   
   
2. Samples & design
   

   
   
3. Build  
   By default, the 2-AFC question will be displayed in even cycles. This means that if you were to add any monadic questions in your 2-AFC test, those questions will cycle through each sample, while the 2-AFC question will display only once, after the second sample or the 4th, etc, if running a back to back test (i.e. in even cycles). To change this, go to the Question options and make the change in the Present question option.
   
   
   You can also evaluate images as samples, and you can allow panelists to indicate that there is no difference between samples.
   
   If you give panelists the option to indicate that there is no difference between samples (found in the Question options), you can place the 'No difference' button either on the right of the samples, or between the samples, or to the left of the samples (found in the Attribute options). At individual attribute level (in the Attribute options), you can specify whether you want to display the 'No difference' button for that attribute.
   
   When this feature is enabled, the question is no longer a 2-Alternative Forced Choice question since panelists are able to select the third option.
   
   When running the reports, you can decide whether the data for 'no difference' will be distributed evenly between samples, or it would be discarded. Please note that the option in the analysis reads 'no preference' because this option is shared with the Preference question data.
   
   
   
   The 2-AFC question is already added to the test along with the 2-AFC related Comment .
   
   
   
   If you wish to ask panelists to evaluate multiple 2-AFC attributes, complete the first attribute setup and clone the attribute. Update the new attribute as necessary. 
   
   
4. Panelists
   
   
   
5. Logistics
   

1. Graphs - Visual representation of the data at the attribute level, compatible with 2-sample tests only. If you ran the test on 4 or more samples, use the PowerPoint report instead.
   
   
2. Data - Raw data displayed in the form of check marks indicating whether the panelists selected a sample as different (as 'more' in the evaluated attribute), or they indicated that there was no difference between samples (if this feature was enabled for them). The Total Correct column provides the count of selections per sample and the 'no difference' button (if enabled in the test).
   
   
3. Reports - Create now or view previously generated reports, or schedule reports to be emailed out at a later time. The following reports are compatible with the 2-AFC question type:
   
1. Standard report - Detailed information about the tabulated results for samples chosen as different (as 'more' in the evaluated attribute), Binomial distribution: Null Hypothesis (Distribution mean, Number of trials, Chance, Distribution variance), Alternative hypothesis (d' (d prime) - available only when 'no difference' option is disabled), the p-value, and the statistical decision.
   
   If a follow up questions were included in the test, information for each follow up question will be available in the report.
   
   
2. Summary report - Summary Statistics by Attribute table outlining the counts for each sample being chosen as different ('more' in an attribute), count of 'no difference' selections if the feature was enabled, the p-value for two-tailed test, p-value for one-tailed tests, and d' (d prime) values for one-tailed tests if 'no difference' was not enabled and if the d' option was selected in the Advanced analysis options screen.
   
   
3. PowerPoint report - Visual representation of the percentages for responses in a PowerPoint Presentation report format.
   
   
4. Panelist completion time report - Provides the average test completion time for the test and the total time for each panelist.
   
   
5. Raw data export - Choose from the Multivariate, Univariate, and Multivariate Columnar raw data export formats to suit your exporting needs.