Statistics

Abstract: An expanded scale, the Near-Death Experience Content (NDE-C) scale, has been developed by Charlotte Martial and colleagues (2020) to assess near-death experiences (NDEs). The scale was modeled after the existing NDE Scale (Greyson, 1983), using reworded scale items, and added five additional items, notably items about a gateway or tunnel and a feeling of non-existence or fear. NDE-C uses a rating scale of 0–4 for each item and thus ranges from 0 to 80. In this article, we raise concerns that several of the reworded items appear to have changed the meaning of NDE Scale item and appear to be significantly more prevalent or less prevalent than the original item, as demonstrated in a Rasch statistical analysis of the NDE-C results. We suggest wording changes for these items so that they return to their original context and meaning. [This page is an abbreviated version of the full article PDF, which can be downloaded here.]

Introduction

Charlotte Martial, University of Liège, BelgiumAn article describing the development of an expanded scale to assess NDEs, "The Near-Death Experience Content (NDE-C) scale: Development and psychometric validation," was published recently in Consciousness and Cognition by researcher Charlotte Martial and colleagues (2020) of the Coma Science Group at the University of Liège in Belgium. Because the current NDE Scale, developed nearly 40 years ago by Bruce Greyson (1983), has been so widely used in NDE research, the Martial et al. article is likely of significant interest to researchers in the field of near-death studies. Greyson is one of the co-authors of the Martial article.

Martial and colleagues followed a three-phased strategy to evaluate the current NDE Scale and develop a revised scale. Several improvements were needed, for example: (a) to include additional content items that are now known to be characteristic of the NDE phenomenon, such as negative emotions experienced in “distressing” NDEs, the decision to come back to the body and physical life, and a gateway or tunnel; (b) to expand the number of responses using a Likert-type response (0 to 4) to each scale item; and (c) to simplify and clarify the wording of the items to make them easier to understand. The result was the Near-Death Experience Content (NDE-C) Scale. The two scales differ in the following ways:

Comparison of the NDE Scales

In both scales, the cutoff score for an NDE represents one standard deviation below the mean.

Of the original NDE Scale items, 15 are included in NDE-C but are reworded ostensibly to be clearer and more specific. The original items #5 (Feeling of peace or pleasantness) and #6 (Feeling of joy) were considered redundant, so #6 was dropped. Conversely, the following five new NDE-C items were added:

  • 15. Feeling of non-existence, of being in a total void, and/or of fear.
  • 17. Made the decision, or was forced, to come back from the experience.
  • 18. Feeling of dying and/or being dead.
  • 19. Saw or entered a gateway (for instance a tunnel or a door).
  • 20. The experience cannot be described adequately in words. 

The research group found that the NDE-C Scale differentiated NDEs from experiences associated with hallucinogenic drugs, meditation, and cognitive trance. However, they found that the scale does not differentiate NDEs that occurred in life-threatening situations from NDE-like experiences (NDLEs) that occurred in non-life-threatening situations such as falling asleep, high anxiety, fainting, or occurring spontaneously. This is the same conclusion that Charland-Verville et al. (2014) reached regarding the original NDE Scale. The features or elements in both NDEs and NDLEs appear to represent a shared common ‘core’ experience. “The NDE-C scale thus aims to identify the content of an NDE, whatever the context in which it has been experienced” (Martial et al., 2020, p. 19).

We found it noteworthy that Bruce Greyson, developer of the original NDE Scale, is a co-author of the Martial et al. (2020) article. His original scale certainly has served the field of near-death studies well—indeed, indispensably. And now his presence among the Martial et al. group seems to indicate his endorsement of the NDE-C Scale.

Commentary by Robert and Suzanne Mays, NDE researchers

Due to its more robust psychometrics, NDE-C would now seem to be researchers’ instrument of choice going forward in the field of near-death studies. However, we would caution that the new scale first needs to be validated with a Rasch analysis equivalent to the analysis done by Lange, Greyson, & Houran (2004). The 15 scale items from the NDE Scale were reworded in NDE-C to be clearer and more specific. Even though the changes were reviewed by external NDE experts and the item wording was revised, the reworded NDE-C items should be validated that they preserve the original meaning from the NDE Scale. In particular, we would expect that the reworded scale items in NDE-C would preserve the item hierarchy consistent with the NDE Scale when administered to an equivalent group of experiencers (NDErs), ideally to the same set of NDErs taking both the NDE Scale and the NDE-C Scale.

Figure 1. NDE Scale hierarchy, with item locations in logits, Lange et al. (2004, p. 167)The Rasch model is a psychometric statistical tool for analyzing responses to questionnaires. In particular, a Rasch analysis provides a measure of the relative “difficulty” of the questions. In the case of NDE scale questions, the analysis provides a measure of the relative prevalence of the scale item responses in the form of a hierarchy, in which the least prevalent items are listed first, down to the most prevalent or common item responses. Lange et al. (2004) applied the Rasch model to a set of NDE Scale responses from 203 NDErs. The item hierarchy for their NDE set is shown here (Figure 1, to the right). In Rasch analysis, the value of an item measure is expressed in a logarithmic metric called logit, which converts the measure to a linear form for ease of comparison of two values. A difference greater than about 0.3 logits is considered statistically significant at p < 0.05.

Rasch analyses of NDE rating scales

Lange et al. asserted that the NDE Scale items form a hierarchy that is preserved “regardless of demographic differences and extreme variation in the intensity of NDE” and that “NDEs indeed appear to form a ‘core’ experience whose basic structure and semantics are preserved.” (p. 173). In our study of Rasch analyses of NDE rating scales, we found that within a set of NDE responses, the hierarchy is indeed preserved, as Lange et al. stated, regardless of demographic differences and variation in NDE intensity.

However, we found that there are hierarchical shifts of one or several pairs of scale items between NDE scale results from different sets of NDE cases. For example, we compiled a set of NDE Scale results from 215 NDErs from the Experience Registry database maintained by the International Association for Near-Death Studies (IANDS, n.d.). The item hierarchy from the IANDS data differed when different subsets of the 215 cases were compared. The differences occurred primarily in the middle section of the hierarchy, from items N15 through N13. In our study, a total of three scale items shifted, namely, N9, N16, and N7. These shifts were due to differences in the specific NDEr responses in the three sets of NDE cases. See Appendix A in the full version of the article for more details of our analysis comparing Rasch hierarchies from subsets of the IANDS database.

We concluded that the Rasch item hierarchy for any set of NDE responses is potentially different. The hierarchy of the middle items can shift based on the mix of the NDEs which are included. However, if the same NDErs are measured on two different NDE rating scales, the Rasch hierarchies of the corresponding scale items should be equivalent. If the item hierarchy from the NDE Scale is preserved in NDE-C for the same NDErs, the two scales can be considered equivalent measurements of the ‘core’ experience. Subsequent NDE-C results can then be compared with earlier NDE Scale results on the basis of the 15 common scale items.

Initial evaluation of NDE-C using estimated hierarchical placement

The order of the Rasch hierarchy of a rating scale can be roughly estimated by the number of NDErs who responded "0" to each item. In our initial analysis, using this rough measurement, we noted several NDE-C items that most likely deviated from the Lange NDE Scale hierarchy.

We studied the structure of NDE-C and how the item wordings from the NDE Scale were adjusted, ostensibly to make them clearer, more specific, more representative, and more relevant. We became concerned that some of the “translated” NDE Scale items seemed to show a different relative prevalence in NDE-C (Mays & Mays, 2020). For example, the NDE Scale item N14 (Encounter mystical presence) is relatively commonly reported in the NDE Scale whereas the corresponding NDE-C item C3 (Heard voices) is quite rare—only a little more common than N3 (Scenes from the past). C3 would therefore appear as the third item in the NDE-C item hierarchy, rather than the ninth item as N14 appears in the Lange NDE hierarchy.

When we checked the rewording of C3 versus N14, we found that the meaning of N14 was significantly altered in the new scale:

  • N14. Did you seem to encounter a mystical being or presence, or hear an unidentifiable voice?
    • 1 = I heard a voice I could not identify.
    • 2 = I encountered a definite being, or a voice clearly of mystical or unearthly origin.
  • C3. You heard one or several voices which did not have any material incarnation.

To us, the essential concept of encountering a “mystical being or presence” in N14 has been lost and has transformed into “hearing voices” in C3. From the published results, C3 is clearly less prevalent in the NDE-C results (having the third highest number of “0” ratings). This means that many NDErs who reported “encountering a mystical presence” on the NDE Scale did not report “hearing voices” on NDE-C, despite there being a 77% overlap of NDErs who were rated on both scales (Martial et al., 2020, p. 10).

This apparent shift of the C3 item’s results is genuinely concerning to us because we would expect that the rewording of the scale item would preserve the meaning of the item in the new scale, and indeed “represent the construct of interest,” as the authors sought. The authors described their process of rewording the scale items:

[S]ubstantial wording modifications were made to the initial items. Indeed, all statements of the NDE-C scale were worded in an affirmative manner. We sought to write clear, unambiguous items in a language that respondents could easily understand and that represent the construct of interest. We were careful not to use a vocabulary that could be considered too vague or difficult to understand. The sentences were written to take into account the great variety of elements experienced in the NDE and to minimize the risk of misunderstanding or misinterpretation. These modifications were based on our experience in the field and on feedback from experiencers about the NDE scale items. (pp. 7–8).

In addition, the authors engaged a panel of three “internationally known NDE experts”:

An appraisal of content validity was performed to assess the NDE-C scale for clarity, specificity, representativity and relevance .... A panel of three internationally known (external) experts from the field of NDEs reviewed and rated the relevance of the complete scale. ... Items were subsequently revised based on external experts’ relevant feedback. (p. 8).

We questioned whether the criterion of “representativity” really meant that the new wording is actually representative of the original meaning of the item. If so, it appears the authors failed to achieve representativity, based on the comparative hierarchical positions—and also on its face—when the meanings of the two items are compared side by side, despite the relevant feedback from the external NDE experts.

It seemed to us that a scale item rarely reported (such as C3 – Hearing voices), which replaces an item commonly reported (in this case, N14 – Encountering a mystical presence) can hardly be claimed to be more specific, more representative, and more relevant than the item it is replacing. If the new item shifts in hierarchical order compared to the existing item, the item is most likely a different item altogether.

Three phases of the Martial et al. project

Martial and colleagues (2020) conducted their development and validation of NDE-C in three phases, each driven by a study (Figure 1, p. 5). The first phase involved the psychometric analysis of the original NDE Scale and Study 1 examined the internal consistency, reliability, concurrent validity, and the factor structure of the scale, based on administering the original scale to 403 NDErs (pp. 5–7).

The second phase involved the development of the new NDE-C Scale and a similar psychometric analysis of NDE-C in Study 2, based on administering the new scale to 161 NDErs (pp. 7–15). There was an overlap of 77% of the NDErs who were administered both scales (p. 10). The third phase (Study 3) involved evaluating the discriminant validity of NDE-C in distinguishing NDEs from experiences associated with hallucinogenic drugs, meditation, and cognitive trance (pp. 15–18).

Rasch analyses of simulated NDE rating scale data

Given our concerns, we did a Rasch analysis of the reported results from the Martial et al. paper (using Supplementary Material G provided by the lead author). We used the Winsteps software (Linacre & Wright, 1998)—level 4.8.0—to estimate the logit values of the scale item measures and the item hierarchies for each scale. Because we did not have access to the individual NDEr ratings for several scale results, we simulated the answers that the NDErs gave based on the total number of answers reported for each scale item (0, 1, and 2 for NDE Scale results and 0, 1, 2, 3, and 4 for NDE-C results).

In a simulation of—for example—NDE Scale results, hypothetical NDEr responses are distributed across the 16 scale items based on two constraints: (a) that each scale item has so many responses of 0, 1, or 2, and (b) that the distribution of NDE Scale scores (7–32) follow the distribution of scores from actual or typical NDE results, for example, 22% score in the range 7–10, 32% score 11–14, etc.

We validated this method of simulating NDEr rating scale responses by running a test comparing the Rasch analysis of the IANDS NDE Scale data (N=215) with a corresponding set of simulated NDEr responses for that data. We found that the Rasch analysis of the simulated responses was nearly identical to the same analysis on the actual NDEr responses. See Appendix B in the full version of the article for further details.

Rasch analysis of the Study 2 NDE-C data compared with Study 1 NDE Scale data

In our Rasch analysis of the Study 1 and Study 2 data, we simulated the NDEr response data as described above and in Appendix B (located in the full version of the article). The results are summarized in Table 2 below, comparing the NDE-C values (N=161) with the corresponding NDE Scale values (N=403). The comparison of these two scale results is particularly valid because 77% of the NDErs who filled out the NDE-C Scale in Study 2 (124 of the 161 participants) also filled out the Study 1 NDE Scale.

The NDE-C item distribution is similar to the NDE Scale distribution in that the items at the top and bottom of the hierarchies appear to have similar relative hierarchical positions. For example, C10–N11 and C13–N3 correspond at the top of the hierarchies; C5–N5, C11–N12, and C1–N1 correspond at the bottom of the hierarchies.

However, in Table 2, we immediately noted that four of the NDE Scale items shifted considerably when placed in the NDE-C hierarchical order, namely: N14-C3, N9-C8, N16-C16, and N10-C9. These apparent hierarchical shifts need to be evaluated from the perspective of the NDE Scale as the reference hierarchy, since (a) the NDE-C has been derived from the NDE Scale and (b) the corresponding NDE-C items are purported to be equivalent to the corresponding NDE Scale items but have greater “clarity, specificity, representativity and relevance” (Martial et al., p. 8).

Study 2 NDE-C results compared to Study 1 NDE Scale results

To evaluate the Study 2 NDE-C results relative to the Study 1 NDE Scale hierarchy, the NDE-C items need to be arranged in the Study 1 hierarchical order. For the purposes of comparison with the NDE Scale results, we can omit the new NDE-C items.

In order to match the NDE-C items that do not initially fit the reference NDE Scale hierarchy, the expected values for those items can be calculated by interpolation based on the proportional difference that is observed in the NDE Scale hierarchy. The details of this process are explained in more detail in the full article.

Figure 4 below shows graphically the results of these calculations. The blue line is the Study 2 NDE-C expected hierarchy. Blue stars indicate interpolated expected values. The green line is the actual NDE-C values with the green diamonds indicating hierarchical shifts greater than ≈0.3 logits. The 95% Confidence Intervals (CI) are shown as error bars. The CI bars for the four green diamonds do not overlap with the corresponding expected value intervals. The red line is the Study 1 NDE Scale reference hierarchy. With the interpolated NDE-C expected values, the shape of the reference hierarchy is mirrored in the blue NDE-C expected hierarchy.

NDE C Expected Hierarchy and Actual Scores

Four of the NDE-C items (C9, C3, C16, and C8) have a large hierarchical shift of about ≈0.3 logits or more, indicated by the green diamonds. These four items are the same item pairs noted in Table 2. If the apparent shift in the item’s value is greater than about 0.3 logits, the difference is significant at p < 0.05 (Lange et al., 2004, p. 166).

Comparing the wording of NDE Scale items with the corresponding NDE-C item

In Study 2, Martial and colleagues administered the NDE-C to 161 participants of whom 124 or 77% had also participated in the Study 1 NDE Scale (Martial et al., 2020, p. 10). For these 124 participants, the only apparent difference between their answers to the NDE Scale and NDE-C was the difference in the wording of the items. If there are significant shifts in the hierarchical placement of NDE-C items relative to the NDE Scale reference hierarchy, any differences in the meaning of the wording of these items should be assessed whether the wording differences account for the hierarchical shifts.

Table 4 lists the different wordings of the four scale item pairs that showed hierarchical shifts greater than ≈0.3 logits which are significant at p < 0.05. We evaluated each pair in turn.

Table 4. Comparison of NDE Scale and NDE-C corresponding item wordings

N10 versus C9: Being aware of things elsewhere

In the NDE Scale, N10 (Aware of things elsewhere) is quite rare—number 3 in the NDE Scale hierarchy—whereas C9 (Aware beyond usual perceptions) is quite prevalent—number 10 in the NDE-C hierarchy when the new items are omitted. The difference in prevalence appears to be due to the context that is set for N10 in the possible answers to the item: 1 = Yes, but the facts have not been checked out; and 2 = Yes, and the facts have been checked out.

Item N10 is asking about perceptions (a) in the physical realm, (b) going on elsewhere (as if by ESP), and (c) that could be checked out. These kinds of perceptions are usually reported by NDErs in the context of being out-of-body within the physical realm and perceiving things going on away from their physical body or when the NDEr is physically unable to see (e.g., under anesthesia with the eyes taped shut). Given that the perceptions were of things in the physical realm, they could later be checked.

This context is missing from the wording of C9, which is much more general. Being “aware of things beyond what your senses usually perceive” can apply to all sorts of perceptions in the transmaterial realm as well as the physical realm, beyond what one’s senses usually perceive, so the large increase in prevalence is quite understandable.

To be more consistent with N10, we suggest the wording of C9 be changed to “You were aware of things elsewhere in the physical realm, that could be checked.” These changes focus the context of the item to perceptions elsewhere in the physical realm that could be checked, that is, that are not ordinarily perceptible but can be verified by checking.

N14 versus C3: Encountering a mystical being or presence versus hearing voices

In the NDE Scale, N14 (Encounter mystical presence) is quite prevalent—number 9 in the NDE Scale hierarchy—whereas C3 (Heard voices without a material incarnation) is quite rare—number 3 in the NDE-C hierarchy when the new items are omitted. Again, the difference in prevalence appears to be due to the context that is set for N14 in “encountering a mystical being or presence or hearing an unidentifiable voice.” Even when there is a voice, it is “clearly of mystical or unearthly origin.”

The context of a mystical being or presence—a definite being—is missing from C3 in hearing “one or several voices which did not have any material incarnation.” In our reading of NDE accounts, voices are nearly always associated with a being or presence, a being who is either directly visible, whose presence is sensed, or who is “heard” as a voice within one’s mind. In other words, voices are associated with a presence. Therefore, item N14 is prevalent among NDErs.

The fact that NDErs recognized “hearing voices which did not have any material incarnation” as rarely applying to their NDE, demonstrates that this wording does not convey the same contextual meaning as N14. To be more consistent with N14, we suggest the wording of C3 be changed to “You encountered a mystical being or presence, or heard an unidentifiable voice.” These changes focus the context of the item back to the experience of a mystical being or presence.

N9 versus C8: Your senses were more vivid versus experiencing unusual sensations

In the NDE Scale, N9 (Senses were more vivid) is prevalent—number 7 in the NDE Scale hierarchy—whereas C8 (Experienced unusual sensations—sight, hearing, smell, touch and/or taste) is less prevalent—number 5 in the NDE-C hierarchy when the new items are omitted. The difference in prevalence with C8 appears to be a completely different meaning from the meaning of N9. NDErs report their senses were more vivid in terms of how real they appeared, frequently using terms like “realer than real,” “hyperreal,” “much more real than everyday life,” and “this physical experience is the dream; the experience of the other side is the reality.” NDErs report that their sight is more acute and easily focused even to far-away objects, colors are much more vivid and vibrant, sounds are heavenly and unlike anything on earth, and so on.

The sense of vivid, hyperreal sensations is completely missing from C8, which lists the possible unusual sensations, rather than the vividness. This difference alone would account for the hierarchical shift, particularly with the overlap of 77% of the NDErs having been administered both scales.

To be more consistent with N9, we suggest the wording of C8 be changed to “You experienced your senses as more vivid than usual.” These changes focus the context of the item back to the vividness of the NDEr’s senses.

N16 versus C16: Coming to a border or point of no return

In the NDE Scale, N16 (Come to point of no return) is prevalent (number 8 in the NDE Scale hierarchy) whereas C16 (Came close to a border or point of no return) is less prevalent (number 7 in the NDE-C hierarchy when the new items are omitted). The difference in prevalence with C16 appears to be due to the context conveyed by the answer choices to N16—coming to a conscious decision to “return” to life; or coming to a barrier one is not permitted to cross, or being “sent back” against my will.

The structure of the original item N16 appears to us actually to be two different questions, (a) coming to a point of no return and (b) deciding to or being forced to return. Apparently Martial and colleagues agreed with this perspective and split N16 into C16 (Came close to a border or point of no return) and new scale item C17 (You made the decision, or were forced, to come back from the experience).

We agree with the choice to split N16. It is likely that the wording and answer choice for N16 has caused many NDErs confusion as to how to answer the item, having to choose between answering about coming to a point of no return and about the decision to return. The comparative prevalence of these three results (least prevalent to most prevalent) is:

  • C16 (Came close to a border or point of no return):
             0.05 logits.
  • C17 (Made the decision, or were forced, to come back):
            -0.05 logits.
  • N16 (Come to point of no return + decision to come back):
            -0.18 logits (the expected hierarchical value).

In our reading of NDE accounts, it seems much more common for the NDEr to be told to return or to decide to return (e.g., “It’s not your time, you must return”) than to come to a border or point of no return. More simply put, N16 offers NDErs more choices to respond to the item. In our view, splitting N16 into C16 and C17 is the correct decision.

Conclusions

In our opinion, the new NDE-C Scale made several improvements over the current NDE Scale (Greyson, 1983).

  • Five new items, C15, C17, C18, C19, and C20, were included in NDE-C, which added several dimensions to the measurement that were absent in the NDE Scale, including: A feeling of non-existence or fear (C15); A gateway or tunnel (C19); and A feeling of dying and/or being dead (C18).
  • NDE-C omitted N6 (A feeling of joy) as redundant with N5/C5 (A feeling of peace or pleasantness).
  • NDE-C split the item N16 (Came to point of no return) into two items, C16 (Came close to a border or point of no return) and C17 (Made the decision, or were forced, to come back).
  • The NDE-C items were reworded as affirmative statements (“You felt …”; “You had the impression …”), rather than as questions. The aim also was to make the items clearer, more specific, more representative, and more relevant.
  • NDE-C used a different Likert-type response scale (0 to 4, rather than 0 to 2), giving more precise and finer-grained scores.

We found that the hierarchical order of different sets of NDE responses were variable, particularly in the middle range of the hierarchy. Our conclusions about the item hierarchy are: (a) the published Lange et al. hierarchical order is not fixed, especially in the middle range of the scale, but is dependent on the particular set of NDE responses; (b) when Rasch measurements are compared between two sets of NDE results, the hierarchy of one set becomes the reference hierarchy and the values of the other scale can be compared to that reference set; and (c) comparing two different scales—such as NDE-C versus the original NDE Scale—can be meaningful only if the same group of participants (NDErs) are measured on both scales.

If the individual NDEr responses for a study are not available, we found that the NDEr responses can be simulated from the overall statistics of the study results, that is, for each scale item, the number of answers of each value (0, 1, 2, 3, or 4). This sort of simulation produces a Rasch measurement hierarchy that is close to the actual hierarchical values.

We simulated the NDEr responses and derived the Rasch scale hierarchy for both the Study 1 NDE Scale (N=403) and Study 2 NDE-C (N=161) results. Since 77% of the NDE-C participants were evaluated on both scales, we would expect that the Study 2 NDE-C hierarchy would closely match the order of the Study 1 NDE Scale hierarchy.

Using interpolated data points, we modeled the “expected item hierarchy” for NDE-C that mirrored the Study 1 NDE Scale hierarchy. Large differences between the expected NDE-C item value and the actual value indicated unexpected hierarchical shifts for several items. In particular, there were apparently statistically significant shifts in three scale items (C9, C3, and C8).

Our analysis of the differences in item wording indicated that the NDE-C wording had altered the original context and meaning of these items. This conclusion strongly suggests returning to the original context and meaning for these items. We have offered our suggestions for wording changes in C9, C3, and C8. In our view, it is important to preserve the item hierarchy from the original NDE Scale, so that subsequent NDE C results can be compared with earlier NDE Scale results on the basis of the 15 common scale items.

Recommendations
  1. Revise the wording of NDE-C items C9, C3, and C8.
  2. Administer both the original NDE Scale and the revised NDE-C Scale—in randomized order—to a large cohort of NDErs.
  3. Calculate Rasch measurement hierarchies for both scale results. Compare the two results with the NDE Scale as the reference “expected NDE-C hierarchy” as shown in Figure 4. We expect that the revised NDE-C hierarchy will closely match the NDE Scale hierarchical sequence for those scale items they have in common, within the 95% Confidence Interval. Because of the split with C17, item C16 will be less prevalent than the “expected” value of item N16 in the “expected NDE-C hierarchy.”

Martial and colleagues “encourage further research involving a Rasch analysis of the NDE-C scale” (p. 19). We think our proposed comparison of the original scale with NDE-C is essential to validate the rewording of the original scale items to the new format.

An important limitation of our conclusions and recommendations is that they are based on simulated data sets. What our results most clearly indicate is the urgency for a Rasch analysis to be conducted on the original data set which, according to standard ethical research practices, Martial and colleagues should still have. If they find similar results as we found, they may find our recommendations helpful.

References

  • Charland-Verville, V., Jourdan, J.-P., Thonnard, M., Ledoux, D., Donneau, A.-F., Quertemont, E., & Laureys, S. (2014). Near-death experiences in non-life-threatening events and coma of different etiologies. Frontiers in Human Neuroscience, 8, 203. https://doi.org/10.3389/fnhum.2014.00203
  • Linacre, J. M., & Wright, B. D. (1998). A user’s guide to Winsteps, Bigsteps, and Ministeps: Rasch model computer programs. Mesa Press.
  • Martial, C., Simon, J., Puttaert, N., Gosseries, O., Charland-Verville, V., Nyssen, A.-S., Greyson, B., Laureys, S., & Cassol, H. (2020). The Near-Death Experience Content (NDE-C) scale: Development and psychometric validation. Consciousness and Cognition, 86, 103049. https://doi.org/10.1016/j.concog.2020.103049