The veridical Near-Death Experience Scale: construction and a first validation with human and artificial raters
Bruce Greyson, Jeffrey D. Long, Janice Miner Holden, Jean‐Pierre Jourdan, Robert A. King, Suzanne Mays, Robert Mays, Titus Rivas, Natasha Tassell-Matamua, Pim van Lommel, Marjorie Woollacott, Patrizio Tressoldi
Frontiers in Psychology October 16, 2025 DOI: 10.3389/fpsyg.2025.1661390 via OpenAlex
Summary
A new scale, the veridical Near-Death Experience Scale (vNDE Scale), was developed to assess how strong the evidence is for perceptions reported during near-death experiences. Thirteen experts used a Delphi method to agree on eight criteria, including timing of investigation, medical conditions, third-party verification, and the number and quality of perceptions, scored on a four-level Likert scale. The scale was then tested on 17 cases of potentially veridical NDEs by 11 human raters and three artificial raters using large language models. In 14 of the 17 cases (82.3%), human and artificial raters agreed at over 75% when considering two adjacent evidence levels, such as moderate plus strong or low plus very low. The scale offers a practical way to evaluate the evidential strength of reported NDE perceptions.
Study at a glance
| Characteristics | Scale development and validation Peer reviewed |
|---|---|
| Population | Near-death experience cases |
| Keywords | Scale ratio Perception Preference Cognitive psychology Social psychology |
| Citations | 3 |
| Key finding | In 14 of 17 cases (82.3%), human and artificial raters showed over 75% agreement on the evidential strength of perceptions reported during near-death experiences when using the vNDE Scale. |
Abstract
Introduction: In this study, we describe the construction of the veridical Near-Death Experience Scale (vNDE Scale), a structured instrument for evaluating the evidential strength of perceptions reported during near-death experiences (NDEs), and its first validation by human and artificial raters. Methods: The construction was implemented using a typical Delphi Method. The first draft of the scale was evaluated by 13 experts in NDE, who were asked to suggest revisions and comments within a month for the first round and 20 days for the second round. Results: A general consensus was achieved on the second round on eight criteria related to the timing of the investigation, the medical and physical conditions, the level of third-person verification, and the number, type, and quality of perceptions reported by the near-death experiencer, to be rated on a four-level Likert scale. The validation phase consisted of the application of the vNDE Scale to 17 cases of potentially veridical NDEs by 11 independent human raters and three artificial raters based on Large-Language Models. In 14 of the17 cases (82.3%), the overall agreement between human and artificial judges was over 75%, considering the two close levels of evidence strength, i.e., moderate plus strong, low plus very low, or vice-versa. Discussion: The vNDE Scale is a practical tool for evaluating the evidential strength of perceptions reported by near-death experiencers.