Kalasin University Journal of Science Technology and Innovation

A Causal Model of Factors Affecting Government Nurses’ Intentions to Use Artificial Intelligence for Counseling in Thailand

Pagamart ONG-ART — 2026-07-02

This study aimed to develop and validate a causal model and to examine the direct and indirect effects of factors influencing nurses’ intention to use artificial intelligence (AI) for counseling. The research instrument was a questionnaire whose content validity was evaluated by three experts. The Item–Objective Congruence (IOC) indices for all items ranged from 0.67 to 1.00. The reliability of the instrument was assessed using Cronbach’s alpha coefficients ranging from .935 to .968. Data were collected through an online questionnaire administered to 396 public-sector nurses selected using a proportionate multistage sampling technique. Descriptive statistics, including mean scores, were used for data analysis. Structural Equation Modeling (SEM) was conducted using LISREL software to test the proposed causal model.

The findings indicated that the proposed causal model fit the empirical data well (χ² = 22.18, df = 19, p = .28, χ²/df = 1.17, RMSEA = .02, RMR = .01, and CFI = 1.00). Regarding total effects, environmental factors exhibited the strongest influence on nurses’ intention to use AI for counseling (TE = .94), followed by personal factors (TE = .13), both of which were statistically significant at the .01 level. In terms of direct effects, environmental factors exerted the greatest direct influence (DE = .85), followed by personal factors (DE = .67), with both effects reaching statistical significance at the .01 level. Regarding indirect effects, environmental factors had the strongest indirect influence through personal factors, with an effect size of .09, which was statistically significant at the .01 level. Together, environmental factors (ENV) and personal factors (PEF) explained 90% of the variance in public-sector nurses’ intention to use AI for counseling (OBI).

Effects of Dried Azolla Meal as a Substitute for Soybean Meal in Diets of Labeo rohita

Apirat Eakchatpanya — 2026-07-02

This study aimed to evaluate the effectiveness of dried Azolla (Azolla pinnata) as a replacement for soybean meal in the diets of rohu fish (Labeo rohita). The experiment consisted of four dietary treatments with three replicates per treatment. Experimental diets were formulated by replacing soybean meal protein with dried Azolla at levels of 0, 20, 40, and 60%, resulting in crude protein contents of 29.42, 29.37, 27.77, and 27.15%, respectively. Rohu juveniles with an average initial weight of 4.91 g were reared for 90 days in cages (2 × 2 × 1 m) at a stocking density of 100 fish per cage. At the end of the experiment, the average weight gain was 20.91 ± 3.76, 19.54 ± 2.69, 21.37 ± 2.74, and 21.95 ± 1.78 g for the 0, 20, 40, and 60% replacement groups, respectively. The average daily gain was 0.23 ± 0.06, 0.20 ± 0.02, 0.28 ± 0.07, and 0.25 ± 0.05 g day⁻¹, while the specific growth rate was 1.82 ± 0.25, 1.70 ± 0.13, 2.01 ± 0.27, and 1.90 ± 0.19% day⁻¹, respectively. Survival rates were 41.00 ± 3.06, 46.67 ± 2.51, 56.00 ± 19.67, and 47.33 ± 4.10%, respectively. Feed conversion ratios were 2.40 ± 0.37, 2.61 ± 0.25, 2.18 ± 0.34, and 2.29 ± 0.25, while feed efficiency values were 0.42 ± 0.06, 0.38 ± 0.03, 0.46 ± 0.07, and 0.43 ± 0.05, respectively. The protein efficiency ratio was 1.36 ± 0.19, 1.31 ± 0.13, 1.67 ± 0.26, and 1.66 ± 0.19, respectively No significant differences were observed among treatments for growth performance, feed utilization, or survival (p > 0.05). However, fish fed the diet containing 40% dried Azolla exhibited a tendency toward higher survival compared with the other treatments without any adverse effects on growth performance. The findings indicate that dried Azolla can be used as an alternative protein source to replace soybean meal in rohu diets at levels up to 40% without negatively affecting growth performance or feed utilization, despite having a lower dietary protein content than some of the other experimental diets.

Response Behavior of a Control System for a Single-Beam Twin-Propeller Vertical Balancing Flight Model Using a P, PI, PD, and PID Controller

Somjate Bunchuen — 2026-07-17

This research aims to develop a single-beam twin-propeller vertical balancing flight model and compare the response behaviors of the system controlled by four controller types: P, PI, PD, and PID controllers. In the experiments, the beam equilibrium position was set to 90°. The response tests were then conducted at reference angles of 110°, 120°, and 130°. The system performance was evaluated in terms of overshoot, settling time, and steady-state error. At the reference angle of 110 °, the PD controller provided the best overshoot performance, with an overshoot of 0.50%, a settling time of 2.01 s, and a steady-state error of 0.55 °. Meanwhile, the PID controller produced an overshoot of 1.86%, a settling time of 2.43 s, and a steady-state error of 1.90 °. At the reference angle of 120 °, the PD controller achieved the lowest overshoot of 0.41%, with a settling time of 4.02 s and a steady-state error of 0.45 °. At the reference angle of 130 °, the PD controller produced an overshoot of 3.73%, a settling time of 1.15 s, and a steady-state error of 4.10 °. The PID controller produced an overshoot of 6.04%, a settling time of 2.97 s, and a steady-state error of 3.20 °. The experimental results show that the PD controller is the most suitable for reducing overshoot and achieving a fast response, while the PID controller can compensate for steady-state error.

Development and method validation for testing formaldehyde in resin

Soonthorn Khwan-On — 2026-07-22

Center of Measurement and Standard Accreditation, Faculty of Science, Prince of Songkla University routinely provides testing specification of synthetic resins. This study aimed to develop and validate an in-house method for determining formaldehyde in resin glue. The method involved distillation of the resin glue to release formaldehyde and reacting it with cyanide which led to a cyanohydrin reaction in alkaline conditions. Then, the amount of formaldehyde was determined by the titration method. The results of the method validation showed that, for the amount of formaldehyde in resin glue, the Limit of Detection (LOD) was 0.31 %w/w. The theoretical and practical limits of quantitation (LOQ) were 0.32 %w/w and 5 %w/w, respectively. By adding formaldehyde standard into resin glue (spiked samples) at low (5 %w/w), medium (15 %w/w), and high (30 %w/w) concentration levels, it was found that the method’s accuracy passed the acceptance criteria at the concentration of 5 %w/w (in the range of 95-102%) and at the concentrations of 15 and 30 %w/w (in the range of 98-101%), respectively. Thus, all accuracy tests passed the acceptance criteria. Likewise, the precision tests were performed by repeating the tests on the same day (Repeatability) and at different days and times (Reproducibility). It was found that all tests passed the acceptance criteria with the Relative Standard Deviation (%RSD) not exceeding 10%. Therefore, the developed method for testing the amount of formaldehyde in resin glue and its method validation can be used as a laboratory method for testing formaldehyde in resin glue and is in accordance with the objectives of the experiment. In the near future, this approach could lead to method development for formaldehyde determination in other resin glues.

Clinical Outcomes and Factors Associated with Hemoglobin A1c Among Patients with Type 2 Diabetes Mellitus at Huai Kapi Subdistrict Health Promoting Hospital

Ampa Seedad — 2026-07-24

This retrospective analytical study aimed to evaluate clinical outcomes and identify factors associated with clinical success among patients with type 2 diabetes mellitus (T2DM) at Huai Kapi Sub-district Health Promoting Hospital, Chonburi Province. Data were retrieved from the medical registration database between January 1, 2023, and February 28, 2025. Patients with a history of at least two follow-up visits were included. The primary outcome was the level of hemoglobin A1c (HbA1c). Paired t-test was utilized to compare the mean HbA1c levels between the two periods, and univariable logistic regression was performed to analyze factors associated with target HbA1c control.

A total of 194 eligible patients with T2DM were included. Most participants were female (74.2%), with a mean age of 66 years. During follow-up from 2023 to 2025, the mean HbA1c changed from 7.9 ± 1.8% to 7.8 ± 1.8%, with no statistically significant difference. When glycemic control was assessed using the target HbA1c level of <6.5%, the proportion of patients achieving the target increased from 21.1% in 2023 to 27.3% in 2025; however, the increase was not statistically significant (OR = 1.4, 95% CI: 0.88–2.24, p = 0.156). Patients receiving at least two antidiabetic medications were 2.1 times more likely to achieve target HbA1c levels than those receiving monotherapy (OR = 2.1, 95% CI: 1.031–4.115, p = 0.041).

In conclusion, after two years of follow-up, neither the mean HbA1c level nor the proportion of patients achieving target glycemic control changed significantly, and the overall rate of glycemic control remained low. These findings indicate the need to improve diabetes care through continuous patient monitoring, and regular treatment adjustment to increase the proportion of patients achieving target HbA1c levels.

Development of Tamarind Seed Shelling Machine Using Factor Analysis with Design of Experiments (DOE)

Akkarapon Namjaidee — 2026-07-27

This research aims to 1) develop a tamarind seed shelling machine and 2) study the factors affecting the efficiency of a tamarind seed shelling machine by using Design of Experiments (DOE). There were 2 sizes of tamarind-seed discharge port designed, 10 and 13 mm., and 2 types of crushing blades, solid blades and perforated blades, with 2 power transmission speeds, 72 and 1,080 rpm. The study of factors affecting the efficiency of a tamarind seed shelling machine by design of experiments was to find the optimization of 3 factors involving seed discharge port size, crushing blade type, and speeds. A full factorial design, 2³ Factorial Design with 2 levels, was used in this study. The results revealed that the optimal operating conditions for a tamarind seed shelling machine were considered factors affecting the largest quantity of shelled tamarind seeds with the shortest time with the solid crushing blades of a seed-discharge-hole size of 13 mm., at a speed of 1,080 rpm. It was deemed the optimal factor level that was able to crush tamarind seeds in an average of 147.30 grams of 18.00 seconds of an average time. The average shelling efficiency was 95.3% with an increase of 7.25%

Development of a Care System for Drug Addict Patients with High Risk of Violence in Community

Narakorn Sareelae — 2026-08-02

This research and development study aimed to develop a care system for patients with substance use disorders at high risk of violence in the community and to evaluate the effectiveness of the developed care system. The study was conducted in three phases: 1) situational assessment, 2) model development, and 3) evaluation. The participants consisted of 70 substance abuse patients at high risk of violence, 70 caregivers, and 26 community health network personnel selected using purposive sampling. The study was conducted from October 1, 2024, to September 30, 2025. Research instruments included the Overt Aggression Scale (OAS), Stress Test-5 (ST-5), drug relapse assessment form, missed appointment assessment form, follow-up evaluation form, focus group discussion guidelines, and in-depth interview guidelines. Data were analyzed using descriptive statistics and inferential statistics at a significance level of 0.05.

The findings revealed that the developed care system for substance abuse patients at high risk of violence in the community, named the “COMMUNITY Model,” consisted of nine components: 1) Community Risk Screening and Early Warning Surveillance, 2) Mobile Outreach and Rapid Home Visits, 3) Continuous Symptom and Behavioral Monitoring, 4) Multidisciplinary Care Team, 5) Use of Google Maps and LINE Group for Patient Monitoring, 6) Family and Community Network Participation, 7) Individual Care Plan, 8) Treatment Continuity and 9) Empowering Families in Patient Care. Finally, the mean post-intervention assessment scores were significantly better than the pre-intervention scores (p < .001)

Development and Performance Evaluation of a Helmet Detection Prototype System Using CIRA CORE

Somjate Bunchuen — 2026-08-04

This research aimed to design and develop a helmet detection prototype system and evaluate its detection performance within Uttaradit Rajabhat University. The system was developed using CIRA CORE and trained with image datasets consisting of 200, 300, 400, and 500 images. The experimental process included image preparation, helmet annotation, and model training using a deep learning training tool. The detection performance of the developed models was subsequently compared with that of the YOLOv3 model. The experimental results revealed that the models trained with 200 and 300 images were unable to detect helmets. The model trained with 400 images achieved an average accuracy of 13.5%, whereas the model trained with 500 images achieved an average accuracy of 57.9%. In comparison, the YOLOv3 model achieved an average accuracy of 48.3%. Furthermore, camera-angle evaluation showed that at the 45-degree angle, the system successfully detected helmets 9 times, achieving an average accuracy of 57.2%, whereas at the 90-degree angle, the system detected helmets 6 times, with an average accuracy of 35.4%. The findings indicate that both the number of training images and the camera angle affect the performance of the helmet detection system.