# Utilizing artificial intelligence to assess academic exam anxiety, perceived stress, and achievement motivation among college students

## TL;DR

This study explores how artificial intelligence can assess college students' exam anxiety, stress, and motivation before and after exams.

## Contribution

The study introduces a fuzzy logic system to evaluate psychological states and validates it against traditional statistical methods.

## Key findings

- Achievement motivation moderates the relationship between stress and exam anxiety only before exams.
- Fuzzy logic system results aligned with traditional statistical findings on exam anxiety and motivation.
- AI tools show potential for early identification of students with high exam anxiety.

## Abstract

Exam anxiety is a multidimensional construct combining physiological reactions and affective responses that can hinder academic performance. Academic stress reflects students’ perceived pressure related to workload, deadlines, and self-evaluation. Achievement motivation refers to students’ drive to attain optimal performance. This study evaluated academic anxiety, perceived stress, and achievement motivation before and after examinations and examined whether artificial intelligence can effectively assess students’ psychological states.

A cross-sectional, repeated-measures design was used. Academic institution students completed an online questionnaire assessing exam anxiety, perceived academic stress, and achievement motivation before and after examinations. Data were analysed using SPSS for statistical modelling. In parallel, a fuzzy logic system (FLS) was developed to model students’ psychological states and estimate exam anxiety and achievement motivation in relation to perceived stress. Outputs from SPSS and FLS were compared to evaluate concordance.

SPSS analysis showed a significant interaction between perceived stress and achievement motivation prior to examinations (b = 0.02, 95% CI: 0.01–0.02, p < 0.001). This moderating effect was not observed after examinations (b = 0.00, 95% CI: −0.01–0.01, p = 0.554). The FLS results were consistent with conventional statistical findings, demonstrating strong agreement in identifying levels of exam anxiety and the role of achievement motivation before exams.

Achievement motivation moderates the relationship between perceived stress and exam anxiety only in the pre-examination period, highlighting the temporal nature of this interaction. The alignment between SPSS and FLS outcomes suggests that artificial intelligence, particularly fuzzy logic systems, can efficiently evaluate students’ academic exam anxiety. These findings support the potential use of AI-based tools for psychological state assessment in educational settings, especially for early identification of students at risk of heightened exam anxiety.