Adoption of Artificial Intelligence-Based Employee Experience Chatbots

Abstract

The Adoption of Artificial Intelligence-Based Employee Experience Chatbots Model (Pillai et al., 2024) was developed as a comprehensive instrument to investigate the factors influencing the adoption intention of AI-based chatbots among employees. This psychometric model is theoretically grounded in behavioral reasoning theory, which provides a framework for understanding how individuals form intentions based on specific reasons for and against a behavior. The scale items were rigorously adapted from established prior research (e.g., Claudy et al., 2015; Gupta & Arora, 2017b; Westaby, 2005b) and subsequently refined through consultation with subject matter experts to ensure robust face validity.

The instrument was empirically validated through a large-scale employee survey, with the resulting data analyzed using PLS-SEM (Partial Least Squares Structural Equation Modeling). The study reported comprehensive results establishing the high reliability and strong validity of the measurement model, supporting its use in assessing technology acceptance within organizational settings.

Keywords

Adoption Intention, Artificial Intelligence Adoption, AI-Based Employee Experience Chatbots, Behavioral Reasoning Theory, Common Method Bias, Face Validity, Perceived Anthropomorphism, Perceived Risk, Technological Anxiety, PLS-SEM, Interactivity, Personalization.

Authors

Pillai, Rajasshrie, Ghanghorkar, Yamini, Sivathanu, Brijesh, Algharabat, Raed, Rana, Nripendra P.

Purpose

The primary purpose of this scale is to systematically assess and quantify employees’ intention to accept and utilize employee experience (EEX) chatbots powered by Artificial Intelligence. The instrument specifically aims to measure the drivers and inhibitors that shape an employee’s decision-making process regarding the integration of these technological tools into their daily work routines.

By capturing these underlying reasons and perceptions, the scale provides organizational researchers and practitioners with critical data necessary for successfully implementing and managing AI adoption strategies within the workforce, ultimately facilitating better human-computer interaction outcomes.

Construct

The scale measures the complex construct of Artificial Intelligence Adoption, specifically focusing on the context of Artificial Intelligence-Based Chatbots used for enhancing the employee experience. The underlying theoretical foundation, behavioral reasoning theory, posits that adoption intention is driven by a balance of reasons for adoption (e.g., perceived intelligence, personalization) and reasons against adoption (e.g., perceived risk, technological anxiety).

Key dimensions explored within the construct include Perceived Anthropomorphism, Interactivity, and the Values of Openness to Change, all of which contribute significantly to an employee’s overall attitude toward accepting or rejecting AI-driven tools.

Validity

The validity of the model was established through multiple psychometric tests, ensuring that the instrument accurately measures the intended constructs. Initial development included expert review to confirm high face validity.

• Convergent Validity: This form of validity was successfully established, as demonstrated by the Average Variance Extracted (AVE) values for all constructs, which exceeded the threshold of 0.50 (Guadagnoli & Velicer, 1988; Hair et al., 2017). This indicates that the items designed to measure a specific construct were strongly correlated with each other.
• Discriminant Validity: Confirmation of discriminant validity was achieved by comparing the squared correlation values between constructs against their corresponding AVE values (Fornell & Larcker, 1981). The study confirmed that the squared variance values were consistently lower than the AVE values, ensuring that each construct measured a unique aspect of the overall adoption model.

Reliability

The scale demonstrated high internal consistency and reliability across all measured constructs. Reliability analysis focused on ensuring the stability and consistency of the items.

• Internal Consistency: The reliability metrics, including Cronbach’s alpha and Composite Reliability (CR) values, were calculated for all constructs. These values consistently exceeded the recommended threshold of 0.8, indicating excellent internal consistency and reliability of the measurement items.

Factor Analysis

A crucial step in the validation process involved analyzing potential threats from method bias to ensure the integrity of the factor structure. The authors specifically tested for Common Method Bias (CMB).

The analysis utilized the single factor Harman’s test (Wang et al., 2018). The results indicated that the single factor variance accounted for only 26.82% of the total variance, which is well below the conventional threshold (typically 50%). This finding strongly suggested that Common Method Bias was not a significant concern in the collected data, lending further confidence to the structural relationships identified in the Structural Equation Modeling.

Instrument

Test Type: Inventory/Questionnaire

Format: Items are rated using a standardized five-point Likert scale, ranging from 1 = “strongly disagree” to 5 = “strongly agree”.

Language Available: English

Population Group: Human (Male and Female)

Age Group: Adulthood (18 years & older)

Population Details: The survey respondents consisted of employees working in the technology sector, specifically those employed by IT (Information Technology), ITeS (IT enabled Services), and E-commerce companies.

Test Methodology: The scale was validated using a robust psychometric methodology including Test Validity, Convergent Validity, Discriminant Validity, Test Reliability, Internal Consistency, Measurement Model analysis, and Structural Equation Modeling (PLS-SEM).

Keywords

Artificial Intelligence, Employee Attitudes, Risk Perception, Anthropomorphism, Organizational and Occupational Measures, Human-Computer Interaction Measures, Technology Acceptance, Chatbots, Behavioral Intention, Likert scale, IT employees.

Authors

Author ORCID Identifier: Algharabat, Raed ORCID: 0000-0002-1870-1708

Affiliation Email addresses: Not provided in source content.

Correspondence Address: Test Location Reference: 2024-44235-001, Table 5, Pages 461-463.

Permissions & Fee and Test Year

The instrument is not commercial and is available for academic use without an associated fee. The scale was developed and validated in 2024.

No file is available for direct download.

Reference’s

• Pillai, R., Ghanghorkar, Y., Sivathanu, B., Algharabat, R., & Rana, N. P. (2024). Adoption of artificial intelligence (AI) based employee experience (EEX) chatbots. Information Technology & People, 37(1), 449–478. https://doi.org/10.1108/ITP-04-2022-0287

Items of the Adoption of Artificial Intelligence-Based Employee Experience Chatbots

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

The test items for the Adoption of Artificial Intelligence-Based Employee Experience Chatbots scale are not publicly available in this record. To obtain the specific test items used in the study, researchers must contact the author or the publisher of the journal, Information Technology & People.