Attitudes Towards Artificial Intelligence at Work Scale (AAAW)

Abstract

The Attitudes Towards Artificial Intelligence at Work Scale (AAAW), developed by Park, Woo, and Kim (2024), is a specialized psychometric instrument designed to quantify workers’ psychological responses and perceptions regarding the integration of artificial intelligence (AI) into the workplace. This 25-item measure assesses six critical dimensions of attitudes towards AI: perceived human likeness of AI, perceived adaptability of AI, perceived quality of AI, AI use anxiety, job insecurity, and personal utility of AI.

The development of the AAAW scale involved adapting items from established literature concerning technology acceptance, human-computer interaction, and AI characteristics. Validation was conducted using three independent samples of working adults, demonstrating robust factor structure, high reliability, and strong validity, positioning it as a valuable tool for organizational psychology research.

Keywords

Artificial intelligence, AI attitudes, workplace psychology, job insecurity, technology acceptance, human-computer interaction, organizational behavior, psychometric scale, Likert scale.

Authors

Park, Jiyoung, Woo, Sang Eun, Kim, JeongJin.

[quads id=5]

Purpose

The primary purpose of the Attitudes Towards Artificial Intelligence at Work Scale (AAAW) is to systematically evaluate the nuanced attitudes employees hold regarding the practical application of AI technologies within their professional settings. The scale moves beyond simple acceptance or rejection, focusing instead on underlying psychological variables that influence interaction and resistance to AI integration.

Specifically, the instrument aims to capture how workers perceive AI’s capabilities (adaptability, quality), its perceived social characteristics (humanlikeness), and the emotional and professional risks associated with its adoption (anxiety and job insecurity). By quantifying these perceptions, the AAAW provides researchers and practitioners with data necessary to predict worker behaviors, manage organizational change, and design effective human-AI collaborative systems.

Construct

The AAAW scale is designed as a multidimensional construct, capturing the complexity of human attitudes towards intelligent automation. It operates on the premise that attitudes towards AI are not monolithic but comprise specific cognitive and affective dimensions relevant to the work context.

The scale measures six distinct psychological dimensions:

• Perceived Humanlikeness of AI: The extent to which workers attribute human cognitive or social characteristics to the AI system.
• Perceived Adaptability of AI: The belief that the AI system can learn, adjust, and perform effectively across various tasks and changing work conditions.
• Perceived Quality of AI: The worker’s assessment of the performance, accuracy, and reliability of the AI output.
• AI Use Anxiety: The level of nervousness or apprehension experienced by workers when anticipating or engaging with AI tools.
• Job Insecurity: The perceived threat to one’s current employment status or future career prospects due to the implementation of AI.
• Personal Utility of AI: The perceived usefulness and benefit of the AI system for the worker’s own productivity, efficiency, or career development.

Validity

The validity of the AAAW scale was established through several rigorous statistical procedures, confirming that the instrument accurately measures the intended constructs. The overall construct validity was confirmed via factor analyses, which supported the hypothesized six-factor structure underlying the scale.

Assessment of convergent validity showed strong results, with the average variance extracted (AVE) values for all six dimensions ranging from 0.52 to 0.77. Since these values exceeded the recommended threshold of 0.50, it confirms that the items within each dimension share a high proportion of variance. Furthermore, discriminant validity was supported by ensuring that all factor correlations were statistically smaller than the square root of the AVE for the respective factors, indicating that the six dimensions measure unique and separable constructs. Finally, criterion-related validity was demonstrated through the scale’s ability to explain unique variance in important organizational outcomes, such as recruiting decisions and employee behaviors related to AI.

Reliability

The reliability analysis confirmed the internal consistency and stability of the 25-item measure across all six dimensions. For internal consistency, the measure utilized Cronbach’s alpha, with scores for all six dimensions exceeding 0.80, which is well above the acceptable standard for psychological scales.

Additionally, the composite reliability (CR) was calculated. The CR values ranged from 0.70 to 0.94. Since all these values met or exceeded the commonly accepted cutoff of 0.70, the scale demonstrated excellent internal reliability, ensuring that the items consistently measure the underlying constructs.

Factor Analysis

The development of the AAAW scale included both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to establish the structural integrity of the six-dimensional model.

The EFA initially identified a six-factor solution that was deemed both interpretable and parsimonious. This initial phase utilized 43 items which were subsequently refined down to the final 25 items based on loading strength and theoretical fit. The subsequent CFA was performed to rigorously test the hypothesized structure. The results for the six-factor model indicated a good fit to the data (χ2 = 673.60, df = 260, CFI = 0.98, NFI = 0.96, TLI = 0.97, RMSEA = 0.04, SRMR = 0.04). Although a second-order factor model (suggesting a general ‘Attitude Towards AI’ factor) also provided adequate fit indices, the researchers concluded that the first-order six-factor model provided the superior and most detailed representation of the data.

Instrument

Test Type: Inventory/Questionnaire

Format: Electronic (Likely administered via online survey platform.)

Language Available: English (Based on the publication language; other translations not specified in the source material.)

Population Group: Working Adults/Employees

Age Group: Adult (Specific age range not specified, but collected from working adults.)

Population Details: Data were collected from three independent samples of working adults across various industries, ensuring a broad representation of employee experiences with technology and the workplace.

Test Methodology: The scale utilizes a 25-item questionnaire employing a 5-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). The methodology included extensive psychometric testing, including construct, convergent, criterion, and discriminant validity analyses, alongside EFA and CFA, to ensure high standards of validity and structural fit.

Keywords

Organizational psychology, AI integration, scale development, psychometrics, job anxiety, humanlikeness, factor validity, CR, AVE.

[quads id=5]

Authors

Author ORCID Identifier: Park, Jiyoung: 0000-0003-4397-9645; Kim, JeongJin: 0000-0002-0562-2488

Affiliation Email addresses: Not specified in the source material.

Correspondence Address: Not specified in the source material.

Permissions & Fee and Test Year

Test Year: 2024 (Based on publication date of the validation study).

Permissions and Fees: The scale was published in an academic journal. Researchers should contact the corresponding author for permission regarding commercial use or large-scale academic replication. Standard academic use may be permitted under fair use guidelines, but official permission is recommended.

Reference’s

Park, J., Woo, S. E., & Kim, J. (2024). Attitudes towards artificial intelligence at work: Scale development and validation. Journal of Occupational and Organizational Psychology, 97(3), 920–951. https://doi.org/10.1111/joop.12502

[quads id=5]

Items of the Attitudes Towards Artificial Intelligence at Work Scale (AAAW)

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

• Number of Items: 25
• Rating Scale: 5-point Likert scale (1 = strongly disagree, 5 = strongly agree)
• Format: Electronic