Knowledge Base

Analysis of Trial Control Data

How to add a Hardware Continuous Variable to an Analysis Profile

1. In the Experiment Explorer, right-click Analysis Profiles and select New, or click an existing profile.
2. Click the Add button next to Hardware continuous.
3. The Hardware continuous variable window opens. From the Device type list, select the hardware device you want to calculate statistics for. For example, if you want to analyze how many pellets your pellet dispenser dropped, select Pellet Dispenser (PTPD-001x).
4. In the Device list the name of the hardware device will appear as you defined it when you assigned the device to the arena.
5. From the Signal list, select which signal you want to analyze. For instance, for the pellet dispenser you can select Number of drops.
6. Click the Trial Statistics tab to choose the statistics for the dependent variable. If you want summarized statistics over all trials, or the groups of trials defined in your Data Profile, click the Group Statistics tab and select additional statistics. Next, click Add.

For details, see Calculate Statistics in the EthoVision XT Help.

Application: Hardware Continuous Variable

By selecting Hardware continuous variable as Dependent Variable, you can calculate, for instance, how many pellets your pellet dispenser dropped or how long your mouse has been drinking from its water bottle. You can also calculate the frequency drinking from the bottle took place. But often many of these drinking occurrences together form a single drinking bout. Analyzing such bouts is not possible in the current version of EthoVision XT. This means that to calculate how often such bouts occurred, you have to export the data to a Hardware log and analyze this in an external program like Excel.

Note: Devices with Fixed Sampling Intervals

Devices such as the Lickometer and PhenoWheel send data to EthoVision XT at fixed times, every two seconds for the Lickometer and every 60 seconds for PhenoWheel. At each sample time, EthoVision XT reads the data since the last reading event:

• For the Lickometer: number and duration of licks.
• For PhenoWheel: number of rotations.

The reading events are not in sync with the trial start and stop. This means that:

• The data of the first reading event after the trial start also include some time before the trial start (less than 2 seconds for Lickometer, less than 60 seconds for PhenoWheel). This time is included in the results.
• The data between the last reading event and the stop of the trial are not included in the results. This time is again less than 2 seconds for Lickometer, less than 60 seconds for PhenoWheel, depending on when the trial stopped.

In most cases the two discrepancies have no effect on your data when the trial lasts at least a few minutes or hours, like in typical PhenoTyper tests.

Tip: If you want the devices to have another sampling rate, contact Noldus.

Hardware State Variable

A variable based on a Hardware continuous variable. You can, for example, define a Hardware state variable Number of dropped pellets is equal to or greater than 1 and calculate, for instance, the duration of this interval.

How to add a Hardware State Variable to an Analysis Profile

1. In the Experiment Explorer, right-click Analysis Profiles and select New, or click an existing profile.
2. Click the Add button next to Hardware state.
3. The Hardware state variable window opens. From the Device type list, select the hardware device you want to calculate statistics for. For example, if you want to analyze the signal from your pellet dispenser, select Pellet Dispenser (PTPD-001x).
4. In the Device list the name of the hardware device will appear as you defined it when you assigned the device to the arena.
5. From the Signal list, select which signal you want to analyze. For instance, for the pellet dispenser you can select Number of drops or In error state.
6. Set the appropriate Value (threshold) or select the value you are interested in. For instance, for Number of drops set a threshold of equal to or greater than 1, for In error state select true or false.
   • Cumulative value: Select this option to calculate/visualize the cumulative value of the signal.
7. Click the Trial Statistics tab to choose the statistics for the dependent variable. Click the Group Statistics tab and select additional statistics if you want summarized statistics over all trials, or the groups of trials defined in your Data Profile. Next, click Add. For details, see Calculate Statistics in the EthoVision XT Help.

Application: Hardware State Variable

You can use Hardware state variables to analyze, for instance, how long a light was on. You can also use Hardware state variables to analyze learning behavior. For instance, when training an animal you may want to know how long it takes before it completes a task (and receives a food pellet). For this purpose, you can calculate the statistic Latency to first for the Hardware state variable Number of drops greater than or equal to 1. For naive animals the latency will be high at the start of the training and will decrease as the animals learn the task in subsequent trials.

Please note that the following question requires a different procedure: What was the average time between "light on" and "lever press" in a learning experiment? It is not possible to define a variable based on two separate hardware events. To answer this question, define a Free interval using hardware events. For more information, see Free interval in the EthoVision XT Help.

Free Intervals Based on Trial Control and Hardware Data

Free intervals are the most flexible solution to analyze slices of time which go from an event, or time, to another event or time. You can use them in the Analysis profile and in the Data profile.

Example: In a DanioVision experiment, analyze the speed of the subjects in the 5-second time interval before the onset of the light stimulus, and in a second interval, in the 5-second time interval after the onset of the light stimulus. Because you want to analyze the speed within the intervals, define those intervals as Free intervals in the Data profile.

How to add a Free Interval to an Analysis Profile

In the Analysis profile, under Custom Variables, click the Add button next to Free interval.

For more information, see Dependent variables in Detail > Custom Variables in the EthoVision XT Help.

How to add a Free Interval to a Data Profile

In the Data profile, under Nesting, click the Add button next to Free interval.

For more information, see Analyze Track Segments > Nesting over a free interval in the EthoVision XT Help.

The Analysis Results

To Calculate the Statistics

1. Click Statistics & Charts in the Project Explorer. The analysis window that appears has the tab Trial Statistics for the analysis results per trial and Group Statistics & Charts with statistics and charts from the summarized results over different trials.
2. If you have different filters, data profiles, or analysis profiles, make your selection on the toolbar.
3. Click Calculate.
4. Optional: Click the Layout button to modify the layout of the results table.

Batch Analysis

To run different analyses with different combinations of filters, do the following:

1. Click Batch.
2. Select the profiles from the lists and click Add.
3. Repeat the previous step for other combinations.
4. Click Calculate.
5. Select the combination of profiles from the lists on the toolbar to view the results.

Export Trial Control Data

You can export Trial Control data just like the other dependent variables.

1. Choose Analysis > Export > Raw Data.
2. Select the following:
   • Track & dependent variables: To export events and intervals defined in the Analysis profile. For example, actions and conditions and intervals in between. This option also includes hardware-based actions and conditions, not the actual communication signals.
   • Hardware log: To export the communication signals between EthoVision XT and the hardware device.
3. Click Start Export.

Exported Raw Data File

In the exported raw data file, each Trial Control event or Trial Control state defined in the Analysis Profile is stored in a column.

• A Trial Control event is exported as 1 (one) in the row corresponding to the time when the event occurs. In the other rows, it is exported as 0 (zero).
• A Trial Control state interval is exported as 1 (one) in the rows corresponding to the time included in the interval. If the interval occurs in multiple instances during the trial, you will see multiple segments of 1.

For more information about exporting track data, see Export tracks in the EthoVision XT Help.

Hardware Log File

When you select Hardware log in the export screen, hardware events are exported to a separate file. The data are organized in columns as follows:

• Time is the time when the signal value is read or the action is taken. Note: This is the time elapsed since the start of the trial, not the start of tracking. For lickometers, values are updated about every 2 seconds.
• Device ID is the name of the physical device that the event refers to.
• Command/Signal is the type of data (Command, or Signal).
• Name is the name of the command or signal.
• Value is the value of the signal at that time.

Each row of the file contains one of the two:

• Rows marked with Signal: The value of the signal since the last time the device was read (for example, the number of licks or the duration of licks since the last reading). Duration of licks is expressed in milliseconds.
• Rows marked with Command: The command (for example, Drop pellet).

• The hardware log file is one per trial, no matter how many arenas and devices are used.
• If you re-do a trial, the log files recorded with the previous instance of the trial are deleted.
• In order to be logged, hardware devices must be connected to the computer and configured in EthoVision XT.
• Notice that the Time in the Hardware log file is the time since the start of the trial, while the time in the Track data file is the time since the start of tracking.

Example: If you have defined a command like Drop pellet as a Trial control event, to find a specific Drop pellet event in the data file, take note of the Time in the Hardware log file for that event and locate that time in the Time trial column in the Track data file.

• For Pellet dispensers: Number of drops is the feedback signal sent by the device to EthoVision. Usually it is recorded 0.04 seconds after the actual Drop pellet command.
• Choose Hardware log if you want to export communication data between EthoVision and the hardware device, even when the device was not used during the trial.
• If you select both Track data and Hardware log, data are exported in separate files when choosing the Text format, and in one file when choosing the Excel format. In the Excel file, Track data and Hardware log data are stored in separate sheets.

Apply Different Protocols to Different Arenas

Introduction

This section is intended for EthoVision users who record data with multiple arenas simultaneously, and want to apply different Trial Control protocols to different arenas.

Example: The camera points at two separate arenas, each with one animal. In arena 1, provide a foot shock 30 seconds from the start of the trial. In arena 2, provide a foot shock at a randomly chosen time between 0 and 90 seconds from the start.

Multiple Arenas in Your EthoVision Experiment

If you work with multiple arenas within an experiment, each arena is treated as an independent replicate. In most cases, you want to apply the same Trial Control protocol to all the arenas. Imagine that you create a Trial Control protocol that stops tracking when the subject has moved a certain distance. You define one Trial Control Settings profile and apply that profile for your trial.

Because subjects differ in their behavior, tracking will stop at different times for each of those arenas. That is because EthoVision XT uses multiple copies of the same Trial Control Settings, one for each arena.

Same Protocol, but Independent Processes

The Trial Control Settings are applied to each arena independently. Suppose that the protocol is "After starting the trial, switch on a light when the animal enters the zone named Food zone". A light device has been installed on top of both arenas, and a zone named Food zone has been defined in the Arena Settings for both arenas.

If the animal in Arena 1 enters its Food zone, the corresponding light will be switched on. However, the light on top of Arena 2 will not be activated, unless the animal in Arena 2 enters its own Food zone too. Although the protocol is the same, it is applied in two copies, one for each arena.

To implement this, select the different physical devices (Light 1 for Arena 1, Light 2 for Arena 2, etc.) on the same line in the Arena - Hardware Mapping window (Arena Settings). This way different arenas refer to the same generic Device name (for example, "Light"). This generic name is then used in the instructions of the Trial Control protocol.

Another typical example is the PhenoTyper setup, where the same protocol is applied to a group of 4 or more cages. Conditions and actions are processed independently for each cage.

Source: EthoVision XT 19 - Trial and Hardware Control Reference Manual