Perception of Psychological Recommendations Generated by Neural Networks by Student Youth (Using ChatGPT as an Example)

The neural network can create recommendations on any topic, including psychological themes. This raises an ethical issue: the generated text may be perceived by the user as expert advice, not being compared with other sources, but still used to make decisions. The goal of the study: assess the perception of text generated by a neural network on standard psychological topics (using ChatGPT as an example) and compare it with the perception of recommendations written by a specialist psychologist. 236 participants took part in the experiment. A quasi-experimental design with one sample was used with the introduction of two equivalent experimental interventions: cases with recommendations written by a psychologist and a neural network. At a significant level, text generated by the neural network was perceived as more attractive and inducing the desire to seek help. There is a significant positive correlation between the positive evaluation of recommendations generated by the neural network with the perception of the significance of electronic resources and the reluctance to seek advice from a real psychologist. A negative correlation was established between the readiness to seek help from a texts were generated by the neural network and the presence of experience in individual psychological counseling.