Determining Individual Intelligence Types and Cognitive Styles Using AI-Based Automated Text Analysis

Abstract

Introduction: This study investigates the application of a novel methodology to evaluate the intelligence profiles and cognitive styles of corporate executives using publicly accessible data. Specifically, we conducted an automated text-based psychological assessment leveraging three AI-driven chatbots and a unique linguistic analysis tool. The outputs from these systems were systematically compared across all evaluation metrics and further benchmarked against manual psychological content analyses performed by the research team. This approach aims to validate the consistency and reliability of AI-powered tools in replicating expert human judgment for leadership trait assessment.

Objectives: The main objective of the work is to evaluate individual intelligence types and cognitive styles using AI-based automated text analysis.

Methods: The tools used for automated text evaluation and analysis included AI chatbots (ChatGPT, Gemini, and Gigachat), a specialized program for psychological text analysis (LIWC), and expert content analysis. The estimates obtained using these tools were verified using rank statistics based on the chi-square criterion.

Results: Based on the study’s results, we can form a cognitive portrait of a top manager. 1) The level of intelligence, erudition, and analytical thinking was assessed as high. 2) The author has a high emotional and social intelligence; social recognition is essential to him. 3) Predominant type of thinking: verbal-logical and abstract-symbolic. 4) The author has a high cultural level. This is confirmed by the complex structure of speech, rich vocabulary, absence of cognitive distortions, and grammatical and logical errors. 5) Originality of thinking can be considered average or low. Many industrial and professional cliches were found in the speech. However, this is a standard situation when a specialist uses them in a relatively formalized finance and business automation area. 6) The manner of speech can be characterized as direct and open rather than secretive and evasive.

Conclusions: Based on the work results, the authors concluded that the level of intelligence and cognitive styles can be measured using automated text analysis tools. Statistical tests yielded reasonably reliable estimates. The data showed a significant positive correlation between human expert analysis and artificial intelligence. In our opinion, the study confirmed the primary hypothesis that the analysis of texts from open sources allows us to form an idea of the critical cognitive characteristics of a person. We also confirmed the hypothesis that chatbots based on artificial intelligence can be used for the psychological analysis of texts to assess a person’s cognitive characteristics.