Introduction

The 3Bird Behavior Logger is a data collection tool for applied behavior analysis. The interface was designed for practitioners conducting treatment while one or more data-collectors record behaviors.


Installation

Windows

After downloading the installer (3Bird Behavior Logger.exe), double-click the file.
win
Click "install". When finished, you should be able to open the application from Start Menu > All Programs.

Mac

After downloading the installer (3Bird Behavior Logger.dmg), double-click the file.
mac
Drag the icon over the Applications folder.


Start Menu

start-menu

  1. Schema List - A list of current schemas saved on your computer.
  2. Create - Opens a new Edit Schema menu for creating a new schema.
  3. Edit - Opens the selected schema into an Edit Schema menu.
  4. Export - Allows the user to export a schema into a ".schema" file, to be shared with other users.
  5. Import - Allows the user to import a new schema by selecting a ".schema" file.
  6. Preferences - Opens the Preferences menu for setting default schema values.
  7. IOA Calculator - Opens the IOA Calculator.
  8. Key-Behavior Mappings - A summary of the selected schema's key-behavior mappings.
  9. New Session Summary - The details to be applied to the upcoming session.
  10. Start New Session - Starts a new session in the Recording menu.

Edit Schema

edit-menu

  1. Client - The identifier of the client.
  2. Project - The research project this schema is for.
  3. Key-Behavior Mappings - The keys used to record target behaviors.
  4. Session Data Directory - The directory session data-files will get saved to.
  5. Session Duration
  6. Add Row - Adds a new row to the Mappings section.
  7. Cancel - Discard all changes and return to Start Menu.
  8. Save - Saves changes and returns to Start Menu. If this is a new schema, it will show up in the schema list.
  9. Delete Schema - Deletes the current schema. Prompts the user for confirmation.

Preferences

preferences

  1. Default Save Location - New schemas will start with this directory.
  2. Filename Format - Since most researchers have a preferred naming structure for their data-files, we allow users to customize them. Check each component that you want to include in the filename. You can also drag-and-drop components to reorder them.
  3. Default Session Duration - New schemas will start with this duration setting. See the edit-schema menu for an explanation of each option.

Recording

recording

  1. Behavior Summary - Each row keeps track of the count/duration of each behavior.
  2. Timer - The timer will increase while recording. If the current schema has a set duration and the timer reaches it, it performs the checked actions as specified in the edit-schema menu.
  3. Start Session/Continue - When pressed, it will start the session timer. You can also press spacebar as a shortcut (see Keyboard Shortcuts). You can only record behaviors while the timer is active.
  4. Add Notes - Displays a window where you can attach notes to the session.
  5. Start New Session - Starts a new session with the same schema and same session-detais. The session #, however, will increment.
  6. Edit Uknown Behaviors - Becomes visible if you have logged an unknown behavior. When clicked, it opens a new menu for assigning values to unknown behaviors.

Logging Behaviors

While a session is playing, press any of the mapped keys to log a behavior. Discrete behaviors flash green and the counter will increment. Continuous behaviors toggle on and off.

Logging New Behaviors

You can also log new behaviors on the fly while recording.

Discrete - press an unmapped key.
Continuous - press an unmapped key while holding shift. Once the new continuous key is added, you don't have to hold shift to toggle it again.

When you pause the session, you can click Edit Unknown Behaviors to assign values to any unknowns you might have logged. If you logged an unknown by mistake, you can ignore it.

Undoing Behaviors

You can undo behaviors using ctrl+z for Windows and cmd+z for Macs (see Keyboard Shortcuts for a full list of shortcuts). There is no "redo" funcationality yet.

Undo works differently depending on what the latest behavior was.
For discrete behaviors it simply removes the behavior and decrements the counter by 1.
For a continuous behavior that has been started but hasn't ended, it reverts the counter back to its position before it started.
For a continuous behavior that has ended, it decrements the counter by whatever the duration was.

Adding Notes to a Session

At any time, you can attach notes to a session. Open the Notes menu by pressing the "Add Notes" button or by using the keyboard shortcut ctrl+n for Windows or cmd+n for Mac (see Keyboard Shortcuts). You can type anything into the text area and your notes will be automatically saved in the session output along with the session's data stream.

Session Output

The app saves logged behaviors in the directory specified by the Schema (see Edit Schema). It outputs two files per session:

The names of these files can be configured in Preferences.

Keyboard Shortcuts

Recording Menu

Session Notes


IOA Calculator

ioa

  1. File 1 / File 2 - Paths to the .raw data-files you wish to compare. You should upload the data output from your behavior recording sessions.
  2. Method - The type of calculation to be done. See IOA Methods.
  3. Block Size/Threshold
  4. Save Options
  5. IOA Summary - After generating IOA results, this box displays the summary of IOA for each key.
  6. Generate IOA File - Calculates IOA between File 1 and File 2 and, if New File is selected, prompts the user to save the result in an Excel file.

IOA Methods

Each method below details how to calculate a percent agreement for each behavior logged by two observers during a session. These methods were derived from two papers (see References).

Exact Agreement

Exact agreement outputs a single percent agreement for a behavior between both observers. The data-logs from File 1 and File 2 get partitioned into intervals of a size specified by block-size. For each interval, the observers are considered in agreement if they both recorded the same number of occurrences within the interval. A behavior's percent agreement is equal to the number of agreements divided by the total number of intervals and multiplied by 100%.

Partial Agreement

Similar to exact agreement, the data-logs from File 1 and File 2 get partitioned according to block-size. A score between 0 and 1 is calcuated per interval by dividing the smaller of the two behavior counts by the larger. If both counts are zero, the score equals 1. A behavior's percent agreement is equal to the sum of all scores divided by the number of intervals, multiplied by 100%.

Time Window

Time window varies from the other two in that it generates two percent agreements for each discrete behavior (one for each observer) and one percent agreement for continuous behaviors.

For discrete behaviors, each occurrence within each data-log is given a score of 0 or 1. A behavor recorded by one observer gets a 1 if the other observer recorded the same behavior within ± threshold seconds. If the behavior does not have a match within the bounds set by threshold, it gets a score of 0. The discrete behavior's percent agreement for each observer is equal to the sum of the scores, divided by the total occurrences, and multipled by 100%.

For continuous behaviors, the threshold does not matter. The data-log is partitioned into 1-second intervals. A continuous behavior's percent agreement is equal to the number of intervals where both observers recorded it, divided by the number of intervals where either of the observers recorded it, multiplied by 100%.


References

MacLean, W.E., Tapp, J.T., Johnson, W.L. (1985). Alternate Methods and Software for Calculating Interobserver Agreement for Continuous Observation Data. Journal of Psychopathology and Behavioral Assessment, 8.

Mudford, O.C., Taylor, S.A., & Martin, N.T. (2009). Continuous Recording and Interobserver Agreement Algorithms Reported in the Journal of Applied Behavior Analysis (1995-2005). Journal of Applied Behavior Analysis, 42, 165-169.*