Self-efficacy and AI
Ways artificial intelligence can support self-efficacy
I’m doing final edits of the Problem Solver before sending it to the printer. This process of editing has taken me to places I didn’t expect. Hyphens for example. Until now I didn’t realise the intricacies involved in their proper use. I’ve just applied them intuitively, which it turns out hasn’t been technically correct.
The main issue has been the term problem solving. I’ve never used a hyphen, whether it be a noun, verb or adjective. However, it should have a hyphen in some cases. None of my reviewers commented on this. But I hate the thought of being wrong, so I revised the whole book to ‘correctly’ apply hyphens. Then I gave it to someone to review.
Feedback: ‘Why does “problem solving” have hyphens in some places and not others?’
Now, what to do? In publications there are examples on both sides. I consulted with my preferred AI assistant to tease out pros and cons and then felt at peace with my decision.
Considering my number one priority is clear communication and language is constantly evolving, I decided to be technically incorrect and remove hyphens whenever referring to problem solving. I’d rather follow reader feedback and risk criticism of the work: choosing readability over perfectionism.
Self-efficacy +1. Sincere apologies to English grammar.
Tina
◼️
Self-efficacy and AI
Belief in our own abilities strongly influences the goals we choose to pursue, how we accomplish those goals and our persistence in the face of challenges and setbacks.
The concept of self-efficacy was established by Stanford psychologist Albert Bandura during the 1970s. It can explain a wide range of behaviours, from how people cope with challenges and react to failure, to their career choices and personal interests.
Bandura identified four primary influences on self-efficacy:
- Previous accomplishments - Successful experiences build belief in our capabilities.
- Vicarious experience - Seeing others like us succeed instils a sense of potential.
- Persuasion - Verbal encouragement from others helps us overcome self-doubt.
- Emotions - How we perceive and interpret our physical and emotional states shapes how we judge our capabilities.
Bandura’s research showed that higher levels of self-efficacy lead to better performance. The concept extends to "collective efficacy," which refers to a group's shared belief in its ability to achieve goals together, impacting social change (Bandura, 1982).
The role of artificial intelligence
People with low self-efficacy are more likely to delegate work to AI rather than complete the subject task and develop associated skills. Conversely, individuals with higher self-efficacy are more likely to take on difficult tasks. Both scenarios lead to self-reinforcing loops in behaviour and outcomes.
Research on the use of generative artificial intelligence tools by university students in Spain found that low self-efficacy and self-control relates to a higher use of artificial intelligence to do academic tasks in the name of the student and to the need to use AI to help interact with others. High self-esteem was associated with lower artificial intelligence use for task completion, lower likelihood of needing AI to support socialisation and a higher preference for high-quality face-to-face interactions (Rodríguez et al, 2024).
It's recommended that educators consider the personal and psychological characteristics of their students when implementing artificial intelligence into their educational practices.
Effectively utilising AI to support self-efficacy depends on an individual’s competencies and goals. These should be the basis for determining the extent and timing of artificial intelligence use for a given task. For instance:
For learning and mastery of skills – AI can support by answering questions on background information, clarification of requirements or identification of methods possible, then provide feedback on the work undertaken by the individual.
For productivity and quality assurance – AI can complement work by delegation or collaboration, doing tasks that have been mastered by the individual, are more effective by computation, are repetitive or low priority. The person reviews outputs, fact checks and revises.
More directly, AI agents can support our self-efficacy through dialogue that aligns with Bandura’s four primary influences: reminding us of previous accomplishments, those of others in our situation, reasons to believe in our abilities and by suggesting ways to overcome emotional or physical stressors.
References
Bandura, A. 1982. Self-efficacy mechanism in human agency. American Psychologist, 37(2), 122–147. https://doi.org/10.1037/0003-066X.37.2.122
Rodríguez-Ruiz, J., Marín-López, I. & Espejo-Siles, R. 2024. “Is artificial intelligence use related to self-control, self-esteem and self-efficacy among university students?” Education and Information Technologies (2024). https://doi.org/10.1007/s10639-024-12906-6