AI Capabilities Scale

Abstract

The Artificial Intelligence Capabilities Scale (AICAP), developed by Abou-Foul, Ruiz-Alba, & López-Tenorio in 2023, is a comprehensive 17-item instrument designed to quantify the capabilities of Artificial Intelligence (AI)-based technologies within organizational settings. The scale broadens the conventional definition of AI capabilities to encompass strategic applications across business model innovation, information management, and computer science.

The development of the AICAP scale adhered rigorously to the prescriptive C-OAR-SE method (Rossiter, 2002), which systematizes construct definition, object classification, attribute classification, rater identification, scale formation, and enumeration. The instrument was validated using a sample of respondents from manufacturing firms located in the United States and the European Union, demonstrating strong psychometric properties through detailed factor analysis, reliability assessments, and validity checks. The authors suggest that the scale is particularly valuable for assessing AI development in complex organizational ecosystems.

Keywords

AI Customer Value Proposition, AI Key Processes Optimization, AI Key Resources Optimization, AI Societal Good, Artificial Intelligence Capabilities, Companies, Organizations, Business Model Innovation.

Authors

Abou-Foul, Mohamad, Ruiz-Alba, Jose L., López-Tenorio, Pablo J.

Purpose

The primary objective of the AICAP scale is to provide researchers and practitioners with a robust and multidimensional tool to assess an organization’s actual capacity to deploy and leverage AI. This assessment moves beyond simple technological adoption, focusing instead on how AI capabilities are integrated across strategic organizational functions.

By expanding the scope of AI capabilities to include areas such as business model innovation and information management, the scale facilitates a deeper understanding of AI’s strategic impact, enabling organizations to benchmark their progress in areas related to customer value, process efficiency, resource management, and societal contributions.

Construct

The construct measured by the AICAP scale is Artificial Intelligence Capabilities, which is conceptualized as a multidimensional construct comprising four distinct subscales relevant to modern enterprise: AI Customer Value Proposition, AI Key Processes Optimization, AI Key Resources Optimization, and AI Societal Good.

This structure reflects the integration of AI across functional boundaries, measuring the capacity of AI-based technologies to generate value. The 17 items were constructed following the C-OAR-SE method to ensure that the scale items accurately reflect the specified theoretical structure and are relevant to organizational contexts, particularly those involving advanced manufacturing and information management.

Validity

The validity of the AICAP scale was established through several rigorous psychometric checks, ensuring both internal consistency and construct separation.

Convergent Validity: Evidence of convergent validity (Convergent Validity) was strongly supported. The Average Variance Extracted (AVE) for all sub-dimensions exceeded the required threshold of 0.50, aligning with the criteria set forth by Bagozzi and Yi (1988). This finding confirms that the items within each subscale successfully measure their intended latent construct.

Discriminant Validity: Discriminant validity (Discriminant Validity) was demonstrated as the average correlations between the latent constructs were consistently below the cutoff value of 0.85 suggested by Henseler et al. (2015). This statistically confirms that the four sub-dimensions of the AICAP scale are unique and distinct from one another.

Reliability

The reliability of the AICAP scale was assessed primarily through internal consistency measures to ensure dependability across the sub-dimensions.

Internal Consistency: All four sub-dimensions of the AICAP instrument exhibited satisfactory levels of internal consistency. The coefficient Cronbach’s alpha ($alpha$) values for every subscale were found to be greater than 0.70, which is the widely accepted standard for demonstrating reliability in new scale development.

Factor Analysis

The theoretical structure of the AI capabilities construct was rigorously examined using sequential factor analysis techniques.

Exploratory Factor Analysis (EFA): Exploratory Factor Analysis (EFA) was conducted utilizing principal axis factoring and ProMax oblique rotation, selecting factors with eigenvalues greater than 1.0. The EFA successfully revealed four main dimensions, providing empirical support for the theoretical conceptualization. Items, with the exception of three, loaded cleanly onto their corresponding factors, with loadings exceeding the recommended cutoff value of 0.60 (Kline, 2014), and no significant cross-loading was reported.

Confirmatory Factor Analysis (CFA): In the Confirmatory Factor Analysis (CFA) stage, three items identified as problematic in a single-factor CFA due to high modification indices were removed from the final set (Bagozzi & Yi, 1988). This refinement resulted in the unidimensional model for each subscale exhibiting acceptable goodness-of-fit indices, finalizing the 17-item structure.

Instrument

Test Type: Original

Format: All 17 items of the AICAP scale are measured using a 5-point Likert scale. The response options are: 1 = Strongly disagree, 2 = Disagree, 3 = Neither agree nor disagree, 4 = Agree, and 5 = Strongly agree.

Language Available: English

Population Group: Human (Male and Female)

Age Group: Adulthood (18 years and older)

Population Details: The scale was evaluated using respondents from manufacturing firms located in the United States and the European Union.

Test Methodology: The methodology involved Test Validity (including Convergent Validity and Discriminant Validity), Test Reliability (including Internal Consistency), and Factor Analysis (including Confirmatory Factor Analysis and Exploratory Factor Analysis).

Keywords

Manufacturing firms, Organizational contexts, C-OAR-SE method, Psychometrics, Internal Consistency, Measurement Model, Servitization, Likert scale.

Authors

Author ORCID Identifier: López-Tenorio, Pablo J.: 0000-0003-4601-0733

Affiliation Email addresses:

• Mohamad Abou-Foul: [email protected] (Al-Azhar University)

• Jose L. Ruiz-Alba: [email protected] (University of Westminster)

• Pablo J. López-Tenorio: No data is Available (UNIE Universidad, Facultad de Ciencias Sociales Aplicadas y de la Comunicación)

Correspondence Address: Mohamad Abou-Foul: Al-Azhar University, Jamal Abdl, Naser St, Gaza, Palestinian Territory, Occupied, [email protected]

Permissions & Fee and Test Year

Permissions: May use for Research/Teaching

Fee: No

Test Year: 2023

Reference’s

Abou-Foul, M., Ruiz-Alba, J. L., & López-Tenorio, P. J. (2023). The impact of artificial intelligence capabilities on servitization: The moderating role of absorptive capacity-A dynamic capabilities perspective. Journal of Business Research, 157, Article 113609. doi.org/10.1016/j.jbusres.2022.113609

Items of the AI Capabilities Scale

IMPORTANT: The following scale items must be preserved in their original language and must not be changed in any way.

The AICAP scale consists of 17 items across four subscales.

Subscales: The measure includes the following subscales:

• AI customer value proposition

• AI key processes optimization

• AI key resources optimization

• AI societal good

AI Customer Value Proposition

• Our company is collecting after-sales insights and uses AI to personalize the customer experience and ensure our customers’ success.

• Our specialized data science team uses tools to calculate our customer’s optimal warranty cost and duration.

• Our company is using machine learning models in pricing and quoting optimization.

• Our company collects and analyzes embedded sensor data to provide our customers with predictive maintenance and operation optimization services.

AI Key Processes Optimization

• Our company is using advanced data science in demand forecasting and stocking.

• Our company is making a strategic data acquisition to fulfill customers’ orders on time.

• Our company integrates AI conversational agents’ capabilities such as chatbots in our next-generation CRM.

• Our company uses advanced robotics and predictive maintenance in our internal operations applications.

• Our company uses intelligence capabilities such as machine vision and edge analytics in enhancing yield optimization.

• Our company uses AI data mining capabilities and big data systems to enhance our product innovation process and bill of material (BOM).

AI Key Resources Optimization

• Our company applies analytics to unified data warehouses to optimize our suppliers’ network.

• Our company uses AI applications to optimize our labor workforce.

• Our company uses advanced analytics to optimize our network’s resources, ensure cybersecurity, and safeguard our data.

• Our company uses AI applications to identify our lowest-cost provider.

AI Societal Good

• Our company trains AI assistants to enhance workplace safety.

• Our company uses applied AI such as deep reinforcement learning to cut our operation’s energy consumption, emission, waste, and equity.

• Our company uses data analytics and benchmarks to provide green solutions to our customers that tackle the most prominent societal challenges such as decarbonization.

Note: Items are rated from 1 = Strongly disagree to 5 = Strongly agree.