Knowledge Base

Adjust Your Selection

• Edit Selection — Click the pencil icon to change your selection.
• Adjust Caption — To give the selected data a more logical or shorter name.
• Remove Data Selection

Compare Data in a Bar Chart

1. Select two or more data sets following the steps in Compare data.
2. Click Bar Chart and select the parameters you want to visualize in your chart.
3. Select what Temporal Aggregation method and Participant Aggregation method you want to use and click OK. Your Bar Chart appears.
4. Hover with your mouse over a filter icon to see what data have been selected.

The Bar Chart in the example below shows the mean intensities of the four negative facial expressions (Sad, Angry, Scared and Disgusted) for two groups of test participants (Male and Female subjects).

Compare Data in a Table and Carry Out a T-Test

1. Select two or more data sets following the steps in Compare data.
2. Click Table and select the parameters you want to include in your test.
3. Set a P-value threshold to carry out a two-tailed t-test. The default value for the threshold is 10%.
4. Select what Temporal Aggregation method you want to use and click OK. Your table appears with the mean values and standard deviation for the parameters that you have selected in step 2. Each row in the table shows the results for one of the data sets that you have selected in step 1.
5. Click one of the rows to test for significant differences. Colored cells indicate significant differences. When you move with your mouse pointer over a colored cell you can see the exact P-value.

The table in the example below shows the mean intensities and standard deviations of four facial expressions of two groups of test participants, Male (upper row) and Female (lower row). The colored cell indicates that the two groups show a significant difference in the intensity of Happy.

Note on Standard Deviations

Each participant is a replicate. So if you have one participant with three analyses, the standard deviations will be zero, because the sample size is one. If you have three participants with two analyses each, the standard deviations will be based on three samples.

IMPORTANT Please note that the appropriate statistical test depends on your data. Your data may not be normally distributed, which means that you need a non parametric test instead of a t-test. Also corrections like Bonferroni on the P-values may be needed for multiple comparisons. Consult your statistician for advice.

Source: FaceReader 10 Reference Manual (Help), Noldus Information Technology