Knowledge Base

Variance

• For most variables, including Turn angle, Angular Velocity and Meander, the square of the standard deviation.
• For the angular variables Heading, Head direction, and Heading to Point, the variance is given by 1 - R/N, where R is the resultant vector length (see the text above for the circular mean) and N the number of samples.

Notes (For All Statistics)

• Not all statistics listed above are available for every dependent variable. To view the list of statistics for a specific variable, double-click its name under Selected Dependent Variables and click the Trial Statistics tab.
• Values of latency and duration can only be a multiple of the sample interval (= 1 second divided by the sample rate).
• To add or remove statistics for a dependent variable, double-click that variable under Selected Dependent Variables and change your selection.

List of Group Statistics

Aim: Group statistics are a summary of trials in a specific experimental group. Here a "group" refers to the tracks selected in a Result box in the Data profile. See Define groups of tracks. Choose between Group mean and 5 number summary.

NOTE: Group statistics are not available when you only select Standard deviation, Standard error or Variance as Trial statistics.

Group Mean

• For most variables, including Turn angle, Angular Velocity and Meander, the Group mean is the arithmetic mean of the trial statistics, calculated for each group.
• For the variables Heading, Head direction, and Heading to Point, the group mean is calculated as the circular mean, with N being the number of trials in the group.

If you select Group mean, the following options are available:

• Standard Deviation. For most variables, the sample standard deviation of the trial statistics, calculated for each group. For the variables Heading, Head direction, and Heading to Point, the group standard deviation is calculated as the circular standard deviation (see above), with N being the number of trials in the group.
• Standard Error. The Standard deviation divided by the square root of the number of tracks used to calculate the Standard deviation. This is also the case for angular variables Heading, Head direction, and Heading to Point.

5 Number Summary

This is the Minimum, the Lower quartile, the Median, the Upper quartile and the Maximum value of trial statistics calculated for each group.

The method used to calculate quartiles is the N + 1 Basis interpolation, which is also used in SPSS. If the sample size is lower than 3, the lower quartile is the same as the minimum and the upper quartile is the same as the maximum value, which is also the same as in SPSS.

NOTE: With 5 number summary you can also plot data in form of Box and whiskers charts. You can also show the original data points for better interpretation. See Group charts > Customize the group charts.

See Also

• List of Trial statistics
• Define groups of tracks
• View the results for groups of tracks

Edit the Analysis Profile

Aim:

• To change the properties and settings of a dependent variable
• To make multiple instances of a dependent variable
• To rename a dependent variable
• To remove a dependent variable from the Analysis profile

After editing the Analysis profile, re-run analysis (Analysis > Results > ...).

To Modify the Properties of a Dependent Variable

You can change the properties of a dependent variable at any time, for example to add one more statistic.

1. Open the Analysis profile that contains the dependent variable you want to modify.
2. Under Selected Dependent Variables, double-click the dependent variable.
3. Edit the variable's properties. For how to modify the variable's properties, choose the variable in Dependent Variables in Detail and see the How to specify section.
4. Choose Analysis > Results > Statistics and Charts, then click the Calculate button.

You can also modify a dependent variable directly in the Statistics result. Right-click the dependent variable header and select Properties. Change the properties of the variable, then click the Calculate button.

To Calculate Multiple Instances of a Dependent Variable

You can add multiple instances of a dependent variable to an Analysis profile. This is handy when you want to compare the effect of different settings on the same dependent variable. For example, two instances of Mobility with different Mobility thresholds.

1. Open the Analysis profile that contains the dependent variable.
2. Right-click the dependent variable under Selected Dependent Variables and select Duplicate. A new row is appended under Selected Dependent Variables.
3. Double-click the new row and set the properties you require for the new instance of the dependent variable.

Additional instances of a dependent variable are given names with a progressive number (for example, Distance moved 2, Distance moved 3, etc.). You can rename those variables (see below).

To Rename a Dependent Variable

1. Open the Analysis profile that contains the dependent variable.
2. Under Selected Dependent Variables, right-click the variable's name and select Rename.
3. Type in the name you want to give and press Enter.

To Remove a Dependent Variable from the Analysis Profile

Under Selected Dependent Variables click the variable's name and press Delete, or right-click the variable's name and select Delete.

Source: EthoVision XT 19 Help, Noldus Information Technology