Learning Efficacy and Memory Feedback

cognition | learning | perception psychology

roles: pilot study experimenter | pavlovia programmer

Background

The strategies that are best for knowledge acquisition are often more painstaking than less effective strategies–‘desirable difficulties’. Can we identify short-cuts to learning that allow us to rely less on strategies that tax our cognitive resources or how about ones that impede it? Reinforcement is one of the principal modes of operant learning and is universally observed. Learning efficacy depends on reinforcing informants giving individuals honest feedback. However, not every informant is credible. Thus, the question becomes:

Experiment Summary

I conducted a three-stage study, comprising a stimulus-response-feedback (SRFB) mapping stage sandwiched between pre- and post-SRFB shape preference task stages on a virtual Zoom experiment. Generally, people can pick up on the credibility of informants quickly and show greater trust in honest informants. They also report liking honest informants more. Negative betrayal by an honest informant decimates any previous trust built-up before the point of duplicity.

The Three-Stage Study Methods

This 17-person study recruited 17-25 year old undergraduate students in an online experiment conducted on Zoom for a course credit compensation. Each subject completed an informed consent form, was de-briefed about the study by the experimenter, and prompted to download the psychology experiment’s program. They each completed a Pre-Preference Task, Learning Task, Feedback and Post-Preference Task to explore if a person’s learning was influenced by the credibility of their informant’s feedback.

Pre-Preference Task

Each subject was asked to indicate their interest on a 6-option Likert Scale in the aesthetics of a randomly-generated face.

Learning Task

After the Preference Task, each subject was tasked to guess which letter matched to a number key on the keyboard. Their goal was to figure out the letter-number pairings mainly through trial-and-error and corrective feedback from the informant. They either indicated a “Good Match” or “Bad Match” to the subject. However, there were two informants where one consistently expressed honest feedback and another was dishonest with how many letters were correctly classified. The credibility identity was unknown to the subject. Thus, the subject also questions their credibility when learning the letter-number pairings with each trial.

Feedback and Post-Preference Task

After the Learning Task, the subject rates the informant’s face on a Like-Dislike Likert Scale depending on their impression of how credible and honest they were when giving the Letter-Number Pairings feedback. They based their preference after evaluating their learning feedback results from the trial blocks in the Learning Task. The feedback screen showed two informants, the Honest and Dishonest version, and the subject’s learning results with how many Letter-Number Pairings they got correct.

Results

In sum, people were quick to learn from and liked honest informants more, but were quick to retract their trust should the credibility of an informant become questionable. This pilot study later helped lay the foundation for the lab’s future studies on positive betrayal (a dishonest informant suddenly becoming honest) and sweet-talking (an always validating informant who confirms all of an individual’s choices but is not necessarily honest). Mainly, establishing best practices with cognitive experimental paradigms delivered in an online setting during the pandemic when labs and data collection were not allowed to operate in-person. This learning and memory study was unfortunately discontinued when the experiment data got lost in a data server fire (crazy, I know — oh well) and lab priorities changed. But if there’s anything you can take away from this research, it’s this — if you want to learn a topic deeply, choose a credible mentor who you can trust will give you honest feedback to improve upon when teaching.