competitions. Epidemiological data on test anxiety in Taiwanese students remain limited. However, globally, the prevalence

of test anxiety across educational stages ranges from 15% to 25%. The severity of test anxiety increases with age, peaking

during high school and junior high school years. Excessive test anxiety in adolescents results in academic underperformance

and poor learning outcomes. In addition, test anxiety leads to cognitive problems, such as negative self-evaluation and

apprehension of underachievement and low self-esteem; physiological problems, such as accelerated heart rate, rapid breathing,

and increased blood pressure; and behavioral problems, such as nail-biting or looking around nervously. Although test anxiety

is common among students, excessive anxiety can reduce their ability to perform effectively in tests, thereby affecting their

academic achievement. Moreover, excessive test anxiety can reduce students’ social functioning and overall well-being. The

aforementioned effects of test anxiety are mediated through reduced concentration. Attentional bias refers to individuals’

tendency to disproportionately focus on highly threatening stimuli during information processing. Empirical studies have

demonstrated that individuals with high levels of anxiety are more likely to focus on high-threat stimuli.

Dot probe task (DPT), a form of attention bias modification (ABM) training, helps reduce attentional bias by training

individuals to shift their focus away from high-threat stimuli. This task frees cognitive resources from threatening stimuli,

thereby promoting emotional regulation and reducing test anxiety in students. However, the therapeutic effects of ABM remain

inconclusive. Moreover, multiple gaps remain in the literature on the effectiveness of DPT in alleviating students’ test anxiety.

First, the effectiveness of DPT is influenced by the intervention location—school or home (remote). Thus, further empirical

research is needed before the widespread application of DPT. Second, whether informal academic tests genuinely elicit test

anxiety–related emotions in students remains unclear. Third, test anxiety has been mainly using self-report questionnaires;

very few studies have included diverse physiological indicators. Finally, most studies have implemented computer-based DPT

programs; computers are less accessible than mobiles. The effectiveness of mobile-based DPT programs remain to be evaluated.

Nevertheless, compared to conventional psychological treatments, computerized and online ABM therapy offers a relatively

cost-effective option, making it worthy of further investigation.

In this study, we developed a mobile-based DPT program for Taiwanese students. This program was developed using test

anxiety–related words as culturally relevant stimuli. It lasted for 6 consecutive weeks. A total of 18 DPT sessions (each lasting

10 min) were conducted between two mid-term tests; the sessions were held thrice a week (Monday, Wednesday, and Friday).

Students’ daily academic tests were integrated into the program to closely simulate real-life test scenarios. Their heart rate, heart rate variability, and systolic and diastolic blood pressure were measured during tests. We hypothesized that the DPT program

would significantly reduce students’ state and trait anxiety, enhance their emotional regulation ability, mitigate attentional

bias, and prevent anxiety-related physiological changes. Physiological indicators were measured 5 days before the mid-term

tests. This study included 13 high school students; of them, 12 students completed the training. Paired t tests revealed that the

DPT program effectively reduced students’ state and trait anxiety, enhanced their emotional regulation ability, and suppressed

(partially) physiological arousal. However, no significant improvement was observed in attentional bias.

To the best of our knowledge, this study is the first empirical exploration of ABM among Taiwanese high school students.

However, the absence of a control group posed a significant limitation to the study’s design, as it restricted our ability to isolate

the true impact of the intervention from other confounding factors. Our findings suggest that the mobile-based DPT program

can help alleviate high school students’ state and trait anxiety and enhance their emotional reappraisal ability. Our study was

exploratory in nature; thus, future studies should expand the sample size and include a control group to enhance statistical

power, improve the representativeness and generalizability of the findings, and offer a reliable basis for evaluating the actual

effects of DPT on test anxiety, thereby allowing for a conclusive evaluation of our hypothesis.

Notably, we recruited students from advanced classes, a group that may exhibit different levels of test anxiety compared

with the general student population. We did not compare intervention efficacy between students with high anxiety levels

and those with low anxiety levels; this limits the broader applicability of our findings, necessitating further analyses for

differentiating between these groups. Furthermore, our program exerted no significant effects on attentional bias; this finding

may be attributable to the insufficient intensity of the stimulus used in the study or the lack of motivation and environmental

pressure during home-based training. Furthermore, different forms of ABM (e.g., stimuli presented as text versus images)

may influence the manifestation of attention bias in different ways (Parsons et al., 2019). Future studies should explore

methods to increase the intensity of stimuli to better induce anxiety responses in participants. Additionally, incorporating game

elements into the intervention design may enhance students’ motivation for home training, thereby improving engagement and

effectiveness.

Our program did not significantly improve students’ physiological measurements, possibly because they were healthy

adolescents, not clinical patients; this made it difficult for us to detect test anxiety–related physiological changes. Furthermore,

the measurements were performed too far from the test dates to accurately capture anxiety-related physiological changes during

actual test scenarios. The extent to which subjective anxiety-related changes correspond to objective physiological changes

remains unclear. To address these limitations and enhance measurement accuracy, future studies should consider using longterm

wearable physiological monitoring devices, which may provide comprehensive data on the physiological effects of ABM.

Overall, the physiological findings from our cohort do not fully support the efficacy of the DPT program in reducing in test

anxiety.

Evaluating the efficacy of ABM training using only changes in attentional bias, self-report scales, and physiological

measurements may pose significant methodological limitations. These methods, while valuable, may not provide a

comprehensive understanding of the underlying mechanisms driving the intervention’s effects. To address these gaps, further

empirical research is needed to substantiate the efficacy of ABM in reducing test anxiety among high school students. Future

studies should also incorporate advanced neuroimaging techniques to assess changes in brain network activity before and

after ABM training. Such approaches can help identify specific neural correlates associated with attentional bias modification,

offering deeper insights into its cognitive and emotional impact.

This study has several limitations. First, as an exploratory study, it involved only 12 participants (3 boys). Thus, caution

must be exercised when generalizing our findings to the broader high school student population. Second, because of the

absence of a control group, the results might have been influenced by the expectancy effect. Future studies should address

these limitations by including control groups, balancing sex proportions for accurate comparisons, and expanding the sample

to include different age groups (to investigate whether DPT exerts varying effects across age ranges). Lastly, although this

study collected both self-reported anxiety levels and physiological indicators, the physiological measurements were only taken

five days before mid-term exams. Future research should consider measuring closer to stress events to more accurately assess

the impact of dot probe task on physiological changes caused by exam stress. This approach will help to better understand the

practical application of dot probe task and enhance its utility in clinical and educational settings.