AI and Tech in Behavioral and Social Sciences

The Effect of Generative AI-Based Digital Marketing Communication Personalization on Customer Trust, Customer Experience, and Repurchase Intention: Evidence from Iranian Online Store Customers

Homa Doroudi ^* iD , Seyedeh Aynaz Jabbari iD , Yasser Heidari iD

With the expansion of e-commerce and the growing role of artificial intelligence in customer interactions, understanding the effects of AI-enabled personalization has become important for online retail performance. This applied, descriptive-survey study examined the effect of generative AI-based personalization of digital marketing communication on customer trust, customer experience, and repurchase intention among customers of Iranian online stores. Data were collected from 384 online customers using a researcher-developed 39-item questionnaire. Content validity was assessed through expert review, sampling adequacy through KMO statistics, reliability through Cronbach's alpha, and structural relationships through structural equation modeling using LISREL 8.8. The results showed that generative AI-based personalization had positive and significant effects on customer trust, customer experience, and repurchase intention. Customer experience and customer trust also showed positive direct associations with repurchase intention, and customer trust significantly strengthened the relationship between customer experience and repurchase intention. The findings suggest that generative AI-based personalization can improve online customer relationships when it is perceived as relevant, trustworthy, and supportive of a positive customer experience.

Media Literacy and Scientific Thinking: The Mediating Role of Science Learning Motivation among Lower Secondary School Students

Vahideh Sadeghi ^* iD , Zahra Tavakkolinezhad iD

Media-rich environments have expanded adolescents’ access to scientific information while also increasing exposure to unsupported claims and misinformation. This exploratory pilot study examined associations among self-reported media literacy, science learning motivation, and scientific-thinking tendencies in 31 students in Grades 7–9. Participants completed an anonymous online 65-item researcher-developed questionnaire comprising media literacy (20 items), scientific-thinking tendencies (25 items), and science learning motivation (20 items), each rated on a five-point scale. Pearson correlations, simple and multiple linear regression, and a 5,000-resample bootstrap analysis were used. Media-literacy scores correlated with science-learning-motivation scores (r = .730) and scientific-thinking scores (r = .799); motivation also correlated with scientific-thinking scores (r = .767), all p < .001. Media literacy predicted motivation (B = .554, β = .730, t = 5.755, p < .001, R² = .533). In the adjusted model, media literacy (B = .392, β = .510, p = .002) and motivation (B = .400, β = .395, p = .013) were associated with scientific-thinking scores (R² = .710). The estimated indirect association was .221 (95% bootstrap CI [.020, .447]). Because the sample was small and non-probability based, all measures were collected concurrently by self-report, and the newly assembled instrument had not undergone full psychometric validation, these estimates should be interpreted as preliminary score associations rather than causal or population-level effects.

The Impact of Artificial Intelligence on Enhancing Customer Experience Through Social Media Marketing: The Mediating Role of Chatbots, Virtual Influencers, and Augmented Reality (Case Study: Digikala Customers)

Mehdi Moslehi iD , Foujan Amiri ^* iD , Mohammad Kheiri iD

The present study aimed to investigate the impact of artificial intelligence on enhancing customer experience through social media marketing, with the mediating roles of chatbots, virtual influencers, and augmented reality. In terms of purpose, this study was applied; in terms of nature, it was descriptive; and in terms of data collection, it employed a cross-sectional survey design. The statistical population consisted of customers of Digikala. Given the unlimited size of the population, the sample size was determined as 384 individuals using Cochran’s formula, and participants were selected through convenience sampling. Data were collected using a standardized 29-item questionnaire based on a five-point Likert scale. Data analysis was conducted using partial least squares structural equation modeling (PLS-SEM). The findings indicated that artificial intelligence had a significant effect on enhancing customer experience through social media marketing (β = .311, t = 10.080). Artificial intelligence also had positive and significant effects on chatbots (β = .360), virtual influencers (β = .524), and augmented reality (β = .616). Chatbots (β = .388), virtual influencers (β = .549), and augmented reality (β = .345) also exerted significant effects on enhanced customer experience. The mediating role of all three variables was confirmed. Model fit was supported by a GOF value of .383 and an R² value of .576 for customer experience.

Developing and Evaluating a Cultural Model of Environmental Behavior in Iran: An Exploratory Sequential Mixed-Methods Study

Mehran Nasr iD , Mojtaba Shahnoushi ^* iD

Environmental crises require not only technical and regulatory responses but also sustained changes in citizens’ values, norms, and everyday practices. This exploratory sequential mixed-methods study developed a culturally grounded model for addressing environmental threats in Iran and conducted a preliminary quantitative evaluation of selected model relationships. In the qualitative phase, academic and professional experts were recruited purposively and through snowball sampling, and semi-structured interviews were analyzed through open, axial, and selective coding. The resulting paradigmatic model identified environmental education, knowledge and awareness, and environmental values as causal conditions; environmental ethics and pro-environmental behavior as contextual conditions; government support as an intervening condition; environmental culture-building and public participation as strategies; and consolidation of environmental culture and effective environmental protection as outcomes. In the quantitative phase, 390 residents of District 12 of Tehran completed a questionnaire derived from the qualitative categories and relevant literature. Partial least squares structural equation modeling indicated positive associations of environmental education, environmental values, environmental knowledge and awareness, public participation, and pro-environmental behavior with the capacity to address environmental threats. The primary endogenous construct had R² = .847 and adjusted R² = .842; communality was .390 and redundancy was .330. Because the design was cross-sectional and complete coefficient-level results were not available for every proposed relationship, the findings should be interpreted as preliminary model support rather than causal confirmation. The study highlights the need to connect environmental education with ethical responsibility, participatory opportunities, and institutional support.

Deep and Ensemble Learning for IoT Network Intrusion Detection: Comparative Performance and Deployment Considerations for AI-Enabled Connected Environments

Sadiq Sahep Majeed iD , Mahmood Deypir ^* iD , Razieh Farazkish iD , Ali Broumandnia iD

Internet of Things (IoT) networks connect devices used in smart homes, education, healthcare, transportation, and critical infrastructure, but their heterogeneity and limited device resources create a broad and rapidly changing attack surface. This study compared four standalone deep-learning architectures—deep neural network (DNN), one-dimensional convolutional neural network (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU)—with all six pairwise probability ensembles formed from these models for binary intrusion detection. The analysis used 461,043 observations and 45 features from the TON-IoT network-traffic data. Data were divided into 322,730 training observations and 138,313 test observations using a 70/30 class-preserving record-level split. Preprocessing included missing-value treatment, categorical encoding, removal of non-informative features, and standardization estimated from the training data and applied to the test data. Models were trained under common optimization settings and evaluated using accuracy, F1 score, area under the receiver operating characteristic curve (AUC-ROC), and confusion-matrix errors. Every pairwise ensemble ranked above the four standalone models by F1 score in this single-split analysis. The DNN + GRU ensemble achieved the highest point estimates, with accuracy of 0.9774, F1 score of 0.9680, and AUC-ROC of 0.9963. It produced 87,926 true negatives, 47,260 true positives, 2,074 false positives, and 1,053 false negatives. The DNN + LSTM ensemble ranked second, with accuracy of 0.9769, F1 score of 0.9673, and AUC-ROC of 0.9957. Because the evaluation used one split and single point estimates, the small differences between leading models should be interpreted descriptively rather than as evidence of statistically significant superiority. For deployment, external validation, calibrated thresholds, explainable alerts, drift monitoring, and analyst oversight remain necessary.

Compression-Resilient Deepfake Detection for AI-Assisted Social Media Content Monitoring: Feature-Level Fusion of ResNet50 and EfficientNet-B0 with a Learnable Visibility Matrix

Samir Qaisar Ajmi iD , Mahmood Deypir ^* iD , Razieh Farazkish iD , Ali Broumandnia iD

Deepfakes circulating through social media create a combined technical, behavioral, and governance challenge because manipulated videos may be redistributed after resizing, re-encoding, and compression. This study evaluated a frame-level detector combining ResNet50 and EfficientNet-B0 through feature-level fusion and an iterative Visibility Matrix procedure. Frames were obtained from FaceForensics++ c23 and Celeb-DF v2 and resized to 128 × 128 pixels. The fused model produced the highest reported point estimates, including 83% validation accuracy, an AUC-ROC of 0.89, balanced accuracy of 0.81, and approximately 18 ms inference time per image. Reported accuracy was approximately 79% on an additional unseen evaluation and approximately 78–80% under a severe-compression condition, compared with approximately 65–67% for ResNet50. These results are descriptive because the available experiment did not report repeated runs, confidence intervals, formal statistical comparisons, or a fully specified external-testing protocol. The detector should therefore be interpreted as a candidate frame-level triage component requiring video-disjoint validation, formal ablation, calibrated thresholds, and human review before operational use.

Identifying and Prioritizing the Dimensions of AI-Enabled Human Resource Empowerment in Government Organizations: Evidence from Semnan Province, Iran

Mohsen Alibeyki iD , Seyed Hossein Hosseini ^* iD , Samaneh Bahramian iD

This study identified and prioritized the dimensions of human resource empowerment in the context of artificial intelligence adoption in government organizations in Semnan Province, Iran. An exploratory sequential mixed-methods design was used. In the qualitative phase, semi-structured interviews were conducted with 20 academic and administrative experts selected purposively. The interview data were analyzed thematically to derive the major dimensions and subdimensions of AI-enabled human resource empowerment. In the quantitative prioritization phase, 20 managers and human resource experts completed pairwise comparisons, and the Analytic Hierarchy Process was used to calculate relative weights. Seven major dimensions and 22 subdimensions were identified. Organizational culture and attitudes received the highest weight (0.268), followed by technological infrastructure (0.249), digital competency development (0.164), digital leadership (0.135), AI-enabled human resource processes (0.086), continuous learning and development (0.062), and strategic governance (0.037). The highest-ranked subdimensions were promoting a culture of innovation and digital transformation (0.144), developing digital and computational infrastructure (0.127), creating integrated organizational data repositories (0.087), training employees in data analytics and artificial intelligence (0.084), and promoting acceptance of AI technologies (0.082). AI-enabled employee empowerment in public organizations depends primarily on organizational readiness, technological foundations, and employee capabilities. The proposed framework should be interpreted as a context-specific prioritization model rather than a validated causal model.

The Role of Digital Human Capital in the Relationship between Smart Technologies and Public Value Creation in Iran's Public Sector

Davoud Hosseinpour iD , Iman Raeesi Vanani iD , Habib Roodsaz iD , Alireza Karimi ^* iD

Digital transformation has become one of the most important trends in public governance over the past decade. Governments are striving to enhance the quality of public services, transparency, and citizen satisfaction by utilizing smart technologies such as artificial intelligence, big data, the Internet of Things, and decision support systems. However, global experiences and findings in Iran indicate that the mere introduction of technology does not guarantee public value creation; rather, the success of smart government depends on the interaction between technology, digital human capital, social trust, and transformational leadership. This study aims to explain the mediating role of digital human capital in the relationship between smart technologies and public value creation, as well as to examine the moderating role of social trust and transformational leadership in this relationship. This article is based on the data, conceptual model, and findings of the dissertation "Designing a Public Value Creation Model in the Public Sector Based on Smart Technology." The results showed that smart technologies lead to public value creation only when they are embedded in the context of digital human capital, social trust, and transformational leadership.

AI and Tech in Behavioral and Social Sciences

About the Journal

Current Issue

Articles

Intelligent Design and Validation of an Administrative Automation Evaluation Model in Government Organizations: An Exploratory Mixed-Methods Study

Designing a Paradigmatic Model for Boxing Development in Iraq: A Grounded Theory Approach and Data-Based Analysis

A Structural Model of Classic Film Noir Within the SLP Framework: A Systematic Content Analysis of the Interaction Among Visual, Narrative, and Character Structures

Challenges in Harmonizing AI-Enabled Smart Contracts with the General Principles of Contract Law in the Iranian Legal System: A Digital Governance Perspective

Designing a Model for the Development of Sport for All Through Modern Technologies in Iraq

An Applied Artificial Intelligence Model for Social Media-Based Natural Crisis Management in Iran

Modeling Marketing Strategies for Small and Medium-Sized Enterprises in the Food Industry Using Reinforcement Learning and Natural Language Processing

Supervised Classification of Civil Unrest-Related Posts in Twitter/X Data: Evidence from the CUT Dataset

Enhancing BPMN Repository Consistency Using Hybrid Similarity Metrics and Scalable Retrieval

Developing a Model of Mass Media Functions for Controlling Corruption in Iranian Sport

Multilingual Fake News Detection Using Cross-Lingual Transformer Models

An Agile Branding Framework for Educational Startups in Volatile E-Learning Markets

Language

Issues

Indexing & Abstracting

Make a Submission

Editorial Board

Journal Statistics Summary

Keywords