Knowledge Base

Definition

A time interval specified by two events of Trial and Hardware Control occurred during the trial. The interval may also occur in two or more instances if the events that mark its start and end occur repeatedly during the trial.

If an interval occurs in more instances during a trial, you can choose to analyze either each occurrence or sum up the results for all occurrences. See the Calculate Statistics Per Interval option below.

How to Specify a Trial Control State

1. Click the Add button next to Trial Control state.
2. Next to From, from the Element list select the Trial and Hardware Control element that makes the criterion for the start of the interval. From the Event list, select the state of that element that makes the start of the interval.
3. Next to To, from the Element list select the Trial and Hardware Control element that makes the criterion for the end of the interval. From the Event list, select the state of that element that makes the end of the interval. Choose which occurrence of the ending event you want to consider.
4. Select Ignore last interval if incomplete to ignore the interval when the Trial Control event that defines the end of the state is not found. If you do not select this option, and the end criterion is not met, EthoVision XT defines an interval up to the end of the trial.

   Example: A Trial Control state is defined "from Trial start to when the condition 'Subject in zone A' becomes true." If the subject never enters the zone, then the end criterion is never met. If you select the Ignore option, the Trial Control state is not defined. If you do not select the option, the Trial Control state is defined from the start to the end of the trial.

5. An interval may occur several times in a trial. If you want to have statistics for each occurrence, select the Calculate statistics per interval option. Next to For consecutive intervals, choose the range of occurrences you want to have in the results.
6. Complete the procedure to add the variable. See Calculate Statistics: Procedure.

Notes

• Do not select Calculate statistics per interval when you want to sum up the results from the occurrences of the state interval in the trial. For example, to calculate the cumulative duration of the state From condition 'In Cue zone' becomes true To condition 'In Feeder zone' becomes true.
• Statistics of duration and latency can only be a multiple of the sample interval (= 1 / sample rate). For example, when you create a condition 'Subject in zone A for >= 3 s', this condition is met when the time elapsed from its activation exceeds 3 s. If the sample rate is 12.5 frames/s (thus the sample interval is 1/12.5 = 0.08 s), the condition is met at the first multiple of 0.08 greater than 3 s, that is 3.04 s. This affects data analysis; for example, the duration of the state 'From condition active to condition true' is 3.04 s.
• The Frequency of a Trial Control state is determined by the start of the state. This means that at the end of a trial, a Trial Control state is counted even if there is no stop event.

Trial Control States Based on Conditions

The sample time at which a Trial Control state starts (or ends) also depends on the statistic used in the condition in the Trial Control rule.

• If you use Current to define the condition (for example, "when Current In zone = true"), the Trial Control state starts (or stops) at the expected sample time. For example, when the animal actually enters the zone (and thus Current = true for that sample).
• If you use any other statistic to define the condition (for example, "when Frequency of In zone >= 1"), consider the two scenarios:
  • When the condition becomes true after the condition box is activated (see How the Trial Control Instructions Are Executed for an explanation of the terms true and active), the Trial Control state starts (or stops) at the expected sample time (see A in the figure below).
  • When the condition is already true when the condition box is activated (for example, a condition "Frequency of In zone = 1" is activated when the animal is already in the zone), such a statistic is only evaluated at the next sample (or in the second next sample, in the case of Heading). In that case, the Trial Control state starts (or stops) one sample (or two) later than expected from the condition (see B in the figure below). However, for the consecutive frequencies of In zone = 2, 3, etc., the condition is already active by definition; therefore the Trial Control state starts at the expected time, when the condition becomes true.

Example: A Trial Control condition has been defined in the Trial Control Settings, which checks that the statistic Frequency of the dependent variable In zone is >= 1 for the Center of the open field. A Trial Control state has been defined in the Analysis profile, which starts when the condition "Frequency of In zone >= 1" is true.

A1: The animal enters the Center zone (and therefore the condition becomes true) after the condition is activated. The Frequency statistic is evaluated at the same sample time. A2: Plot of the variables In zone for the Center and the Trial Control state. The Trial Control state starts at the expected sample time, that is, when the animal actually enters the zone.

B1: The animal is already in the zone when the condition is activated. Therefore, the condition becomes true at the same time that it is activated, but the Frequency statistic is evaluated at the next sample (here, sample rate = 5/s, thus after 0.2 s). B2: Same plot as A2; here, the Trial Control state starts 0.2 s after the In zone state.

See Also

• The EthoVision XT 19 - Trial and Hardware Control Reference Manual. To open this manual, in the Apps screen under Noldus choose EthoVision XT 19 Other Documentation.

Application

You can use Trial Control states to test whether trial control works as expected, and for analyzing learning behavior. For example, calculate the duration of the Trial Control state 'From Cue light ON to Subject in Feeder zone' to see whether this interval decreases during a trial.

Source: EthoVision XT 19 Help, Noldus Information Technology