MIREJ: Multidisciplinary Innovation Research Journal

A Comparative Evaluation of Large Language Models for Enterprise Deployment: Performance, Safety, Cost, and Scalability Using a Multi-Criteria Decision Framework

Erarda Vuka, Jurgen Mecaj — Tue, 28 Jul 2026 00:00:00 +0700

The proliferation of large language models (LLMs) across enterprise, research, and public-sector applications has created an urgent need for rigorous, multi-dimensional evaluation frameworks. This paper presents a comprehensive comparative analysis of seven state-of-the-art LLMs — GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro and Flash, LLaMA 3 70B, Mistral Large, and Claude 3 Haiku — across eight evaluation dimensions: benchmark accuracy, safety alignment, cost efficiency, inference latency, context handling, deployment flexibility, multilingual capability, and scalability. A weighted Multi-Criteria Decision Analysis (MCDA) framework is applied to produce transparent composite rankings from empirical benchmark data using five standardized benchmarks (MMLU, HumanEval, HellaSwag, GSM8K, MATH). Results indicate that Claude 3.5 Sonnet achieves the highest MCDA composite score (0.801), driven by accuracy (90.4% MMLU, 92.0% HumanEval) and safety alignment (4.9/5). Gemini 1.5 Flash emerges as optimal for cost-sensitive deployments ($0.075/1M tokens; 210 tok/s). The paper analyzes architectural trade-offs between dense transformers and Mixture-of-Experts designs, provides a deployment recommendation matrix, and contributes an extensible, evidence-based decision framework for enterprise AI practitioners.

Social-Emotional Skills of Elementary School Students: A Comparative Analysis Based on Ability Levels (HAS vs STD) in Rural and Urban Areas

Tue, 28 Jul 2026 00:00:00 +0700

Student development in elementary school is not based solely on academic ability but also on social-emotional skills (SES), which support self-regulation, peer relationships, and adaptation to school demands. Comparative evidence on how SES varies according to cognitive ability level and place of residence remains limited. This study analyzes differences in SES among elementary school students based on ability level (High-Ability Students/HAS vs. Standard Students/STD) and residential context (rural vs. urban), including the interaction between the two. A comparative quantitative approach with a 2x2 factorial design was used. Data were collected from 240 fourth- and fifth-grade students (103 HAS, 137 STD) using the Fair Cultural Intelligence Test (CFIT), a sociodemographic questionnaire, and the EACSE-CA scale, which consists of five dimensions of SES. The data were analyzed using descriptive statistics, Cramer’s V, Levene’s test, MANOVA with Pillai’s Trace, and univariate F-tests. Ability level significantly affected all five SES dimensions (F = 4.255, p = 0.001, partial eta squared = 0.084), while residence and the ability-residence interaction were not significant. HAS students scored higher than STD students in basic skills, interpersonal relationships, problem-solving, and emotional regulation, with no difference in goal-setting. SES differences are driven more by cognitive ability level than by residential area. The study enriches the empirical literature on elementary students’ SES through a novel comparative design integrating CFIT and sociodemographic data, offering a foundation for adaptive, ability-sensitive social-emotional learning strategies applicable across both rural and urban schools.

The Influence of Intelligence Quotient and Study Habits on the Academic Achievement of Elementary School Students

Muhamad Ali Nurdin, Dimas Yanuar, Rio Andriana, Betari Indira Maharani — Tue, 28 Jul 2026 00:00:00 +0700

Elementary school students’ academic achievement remains a significant concern in Indonesia. The 2022 Computer-Based National Assessment indicated that many elementary school students in West Java had not yet achieved minimum competency standards in literacy and numeracy. Academic achievement is influenced by various internal factors, including cognitive ability, represented by Intelligence Quotient (IQ), and non-cognitive factors such as study habits. Although both variables have been widely examined in Western and East Asian contexts, studies investigating their simultaneous influence among Indonesian elementary school students, particularly in Cirebon City, remain limited. This study aimed to analyze the simultaneous effects of IQ and study habits on the academic achievement of students in Grades IV–VI in Cirebon City. A quantitative approach with an explanatory survey design was employed. The participants were 240 students aged 9–13 years, selected through proportionate stratified random sampling. IQ was measured using the Culture Fair Intelligence Test, while study habits were assessed using a validated questionnaire. The data were analyzed using multiple regression. The results showed that IQ (β = 0.485, p < 0.001) and study habits (β = 0.151, p = 0.008) had significant partial effects on academic achievement. Simultaneously, both variables explained 24.7% of the variance in academic achievement (R² = 0.247, F = 38.901, p < 0.001). IQ was the stronger predictor. These findings highlight the importance of strengthening both students’ cognitive abilities and study habits to improve academic achievement in elementary education.

The Role of Motivational Climate in Predicting Emotional Intelligence and Self-Concept Among Elementary School Students

Tue, 28 Jul 2026 00:00:00 +0700

This study examined the relationship between motivational climate (task-involving and ego-involving climate), emotional intelligence, and self-concept in 280 students in grades 5–6 of public elementary schools in Cirebon City, West Java, using a cross-sectional descriptive-explanatory design and PLS-SEM analysis (SmartPLS 3). The results showed that task-involving climate had a positive and significant effect on ego-involving climate, emotional intelligence, and self-concept, as well as emotional intelligence had a positive and significant effect on self-concept; On the other hand, ego-involving climate had no significant effect on these two variables. In addition to extending the application of Achievement Goal Theory to the general learning context in elementary schools and giving teachers a useful foundation for creating a classroom environment that prioritizes effort and learning processes, these findings confirm task-involving climate as the primary factor influencing students' emotional intelligence and self-concept.

Analysis of the Influence of the Missouri Mathematics Project Learning Model on Elementary Students' Mathematical Connection Skills and Habits of Mind

Khofifah Amaliyatul Maula, Ananta Jelita Putri Ayu — Tue, 28 Jul 2026 00:00:00 +0700

This study examined the effect of the Missouri Mathematics Project (MMP) on elementary school students' mathematical connection ability and habits of mind. A quantitative study employing a quasi-experimental pretest-posttest control group design involved fifth-grade elementary school students. Data were collected using a mathematical connection test and a habits of mind questionnaire. The data were analyzed using descriptive statistics and Multivariate Analysis of Variance (MANOVA) after satisfying the assumptions of normality and homogeneity. The findings revealed that students who learned through MMP obtained higher mean scores in both mathematical connection ability and habits of mind than those who received conventional instruction. The MANOVA results indicated that the differences were not statistically significant for mathematical connection ability (p = 0.314) or habits of mind (p = 0.110). These findings suggest that although the structured learning stages of MMP encouraged students to engage in problem-solving activities, review prior knowledge, and practice mathematical procedures systematically, the intervention duration was insufficient to produce significant improvements in students' mathematical connection ability and thinking dispositions. The value of this study lies in providing empirical evidence that the effectiveness of MMP should be interpreted by considering contextual factors, including students' initial mathematical competence, instructional duration, and classroom learning conditions, rather than relying solely on score improvement. The study concludes that MMP has the potential to support the development of mathematical connections and productive thinking habits; however, longer implementation periods and complementary instructional strategies are required to achieve statistically significant learning gains in elementary mathematics education.