OHAIRE coding system

Behavior coding for Human-Animal Interaction Research


The following guidelines will maximize the use of the OHAIRE-v3 and the range of behaviors it can capture in naturalistic settings.


Children should be video-recorded continuously during unstructured interaction opportunities. The coding system is intended to capture naturalistic behavior, as it occurs in the absence of scheduled or structured activities. However, if there are multiple participants or conditions, it is recommended that sessions be standardized and consistent to allow for comparisons across conditions or over time. Available materials should give participants opportunities to engage in both social and isolated behaviors. For example, providing participants a variety of toys that can accommodate both group and solitary play. Activities that only encourage solitary behaviors, such as reading quietly, do not allow for coding social behaviors. Likewise, interventions that prompt participants to engage only in social behaviors do not allow for coding isolated behaviors.

Social Communication

In order to code social communication behaviors, there should be at least one other person in the frame of the video. In many cases, activities should include at least one adult and one peer if the research question addresses social communication behaviors with partners of different age cohorts. The person behind the video camera should not initiate communication with participants, as the OHAIRE is not designed to code this type of interaction.

Interactions with Animals and Objects

Study Design
Depending on your research questions, you may choose to code the behavior of participants during the intervention, or at another time (e.g., before and after the intervention). Choosing to examine the behavior during the intervention will provide insight into the effects of the interactions taking place between the participant, his or her social environment, and the animals or experimental objects. Choosing to code the behavior outside the intervention enables blinding raters to the experimental and control conditions and explores the generalizability of the effects of the intervention. Neither design is superior, but each is designed to answer specific questions.

Any animal species can be included for coding with the OHAIRE. To date, the OHAIRE has been used to code the outcomes of animal-assisted activities with guinea pigs, horses, and dogs.

We recommend including a control condition that incorporates similar activities to the ones with animals. For example, to control for the novelty of the animal, it may be suitable to include experimental objects that are also novel to participants.
Control condition experimental objects that have been used in studies previously coded with the OHAIRE include a set of toys or one interesting toy (e.g., marble/ball ramp toy), a life size stuffed horse, and a stuffed dog.

Video Recording

The image quality should allow coders to discern participants' facial expressions, such as smiling or frowning, as well as the direction and target of eye contact. The video sample should include a wide enough angle to include the environment and all potential social interaction partners. The recording of a participant’s perspective (e.g., a camera mounted on a child’s helmet) should not be used, as this would limit the ability to view the participant’s facial expressions in response to the environment.

Recording sound quality should be high enough to hear vocalizations including cries, talking, and the content of conversations during playback.

Recording materials
To ensure data integrity, we recommend periodic checks of the output data files. Data files should be saved in a video format (typically .mp4 or .mov), and demonstrate sufficient image and sound quality. Additionally, we recommend checking the battery life and the memory usage of the recording device before each day of filming.

File organization

File naming
To ease the identification and selection of video files for coding, these should be organized and named consistently. File naming can refer to the condition, participant, or time of the recorded session. We will subsequently blind our coders to this information in order to reduce bias.

Participant list
To facilitate the identification of participants during video coding, we recommend taking a photograph of each participant and collating the pictures in a folder with participants’ names or study identification numbers.