In this study, we introduced a novel setup to conduct naturalistic laboratory experiments where participants pursue and interact with physically and socially present actors. So we aim to understand the mechanisms of action perception in real world settings. A significant challenge in experimental research is to design a naturalistic experiment without losing control over the stimuli at the timing.
Although there are some tools such as shutter glasses, they do not provide an opportunity to present real and digital stimuli on the same media. Traditionally, visual perception and cognitive neuroscience research study action perception through tooling displays, for example, images or videos. However, actions in the real world are performed by physically present actors, happen here and now, and are actable.
Therefore, with this setup, we aim to conduct studies that have high ecological validity. The unique aspect of our setup is a transparent OLED screen to which the participants watch the real actions of a real actor. While the timing of their presentation is precisely controlled.
The screen can become opaque and be used as a 2D display when the actor needs to be invisible. We believe that our setup will help psychologists, neuroscientists, and social interaction researchers to reveal fundamental and previously inaccessible cognitive and neural mechanisms of action perception. It will also help bridge the gap between traditional laboratory experiments, which have a high degree of experimental control, and the studies which conducted in entirely unconstrained and natural settings.
To begin, open the experiment software on the experiment PC, and run the ExperimentScript1. m script. Enter the participant ID and age.
Then, the script initiates the experiment displaying across at the center of the participant display, and a warning about pressing escape button to start part one when the participant is ready. Open the experiment software on the actor PC and open the ActorMachine. m script, but do not run it yet.
Place the camera PC near the experimenter PC and ensure that the security camera footage is not visible to the participant. After ensuring that the two actors are ready in their cabinet, draw the curtains and straighten them from the participant area, and turn off the backstage lights. Next, welcome the participant to the main experiment room and direct them to sit in front of the participant display.
Instruct the participant to position themselves such that the cross is in the middle and straight ahead. Give a brief overview of the experiment by referring to the explanations and durations displayed on the whiteboard. Turn off the lights in the participant area.
Before starting experiment part one, inform the participant to complete lexical training. When the participant says that they are ready for the first part, press the escape button. When the participant finished reading the instructions, start the training by matching the target concepts with the attributes.
After the participant completes the training, click on the arrow icon to exit the first part. Inform the participant that lexical training is over and the experimenter is taking control of the mouse to start the second part of the experiment. Next, for experiment part two, open the ExperimentScript2.
m script and wait for the prompt waiting for the actor PC.Then, ring the bell to signal the actor to run the script on the actor PC to enable the connection with the experimenter PC and wait for the prompt, Experiment Part 2 is ready. Inform the participant that the screen will become transparent while they watch short actions and ask them to say out loud what each action is. Press the escape button when the participant says they are ready for the second part.
After the participant watches each action, wait for a response from the participant, then listen to the answer of the participant and write them. Type the participant's response about the action in the dialogue box. Type 1 for correct identification or 0 for wrong identification of action in the second dialogue box.
Watch the backstage from the security camera footage on the camera PC.To start the identification for actor two, wait for the actor to change. When one of the actors waves their hands toward the security camera with the I Am Ready"gesture, press the escape button. When the participant sees the identification is complete warning and exits the part by clicking on the arrow, inform the participant that the action identification is over and the experimenter is taking control of the mouse to start the third part of the experiment.
Next, for experiment part three, open the ExperimentScript3. m script. Inform the participant that they will watch the actions of both actors, and then they will click on the option they think is suitable.
When the participant is ready, press the escape button. Check whether everything goes well backstage and confirm the right actor is conducting the right action during the blocks. After replacing the actors, when actor one waves their hands with We Are Ready"gesture, press the escape button to start the next block.
The participant sees one action through the participant display in transparent mode and evaluates the action by selecting one of the choices at the corners of the participant display in its opaque mode. Once the participant sees the, The experiment is over and thank you"prompts, thank the participant, complete the debriefing, and obtain signatures before sending the participant out of the experiment room. The response times of the participants responding high to the actions of action class one in the agency dimension were significantly higher than for the actions belonging to action class two.
However, these longer response times were not observed in the MD and AUC measurements. While responding low to action class one in the experience dimension, the participants spent significantly more time than they spent for action class two, and this was also apparent in the MD and AUC scores. A significant effect of the action class was observed in the MD and AUC scores of low answers in the agency dimension.
The MDs of the mouse trajectories toward low responses were significantly higher, and the AUCs of the mouse trajectories were significantly larger while the participants were evaluating actions belonging to action class one compared to action class two. Further, participants hesitated toward the high alternative and moved toward low response more when they were evaluating actions from action class one compared to action class two.
