As psychology evolves, the phenomenon of concept overlap becomes more pronounced, increasing participant burden and complicate data interpretation. This study introduces an Embedding-based Semantic Analysis Approach (ESAA) for detecting redundancy in psychological concepts, which are operationalized through their respective scales, using natural language processing techniques. ESAA utilizes OpenAI’s GPT-3 large model to generate high-dimensional semantic embeddings of scale items and applies hierarchical clustering to group semantically similar items, uncovering potential redundancy. In three preliminary experiments, ESAA was tested on well-known psychological scales, such as Conscientiousness, Gratitude, and Grit. The experiments assessed ESAA’s ability to (1) converge semantically similar items, (2) discriminate semantically distinct items, and (3) identify overlapping scales measuring concepts known for redundancy. Additionally, comparative analyses were conducted to assess ESAA's robustness and incremental validity against the most advanced chat bots based on GPT-4. The results demonstrated that ESAA consistently produced stable outcomes and surpassed all evaluated chatbots in performance. As a novel, objective approach for analyzing relationships between concepts operationalized as scales, ESAA has potential to facilitate future research on theory refinement and scale optimization.