Chapter 7

Good Evidence and Bad Evidence

How to Tell a Strong Case from a Weak One

Two Stories, Two Kinds of Evidence

Picture two people sitting across from you. Both nearly died. Both say they had an experience on the other side. The first one tells you, “I felt a deep peace and a great love. It was the most real thing I have ever known.” The second one says, “I floated above the operating table and watched the surgeon use a strange tool that looked like a small electric saw. I described it to him afterward, and he confirmed every detail.”

Both of these accounts matter. Both deserve to be heard with respect. But they are not the same kind of evidence. The first is a report of inner feelings. We have no way to check whether the peace was real, except to trust the person telling us. The second is a report of something that happened in the physical world, something other people could look into and either confirm or deny.

That difference is what this chapter is about.

If we want to know whether near-death experiences give us real evidence that the soul survives death, we need to know what counts as strong evidence and what counts as weak evidence. We need a way to sort and weigh these stories. Not because feelings do not matter. They do. But feelings alone cannot settle the question of whether a mind was actually working apart from a dying brain. For that, we need facts that someone else can check.¹

Think about it this way. If you were on a jury, and a witness said, “I feel in my heart that the defendant is guilty,” you would not convict on that alone. But if the witness said, “I saw him walk out the door at 9:14 p.m., and the security camera confirms it,” that is a different kind of testimony. Both witnesses may be honest. But only one of them has given you something you can check.

I learned this the hard way. When I began my doctoral research on near-death experiences, I read hundreds of accounts. Many of them moved me deeply. But when I tried to sort them by the strength of their evidence, I quickly found that most of them, however sincere, gave me nothing I could verify. A handful, though, gave me details that had been checked by a nurse, a doctor, or a family member. Those were the ones that changed my thinking. Those were the ones I could build on.

In this chapter I want to hand you the same tools I used. By the end, you will know what makes an NDE account strong, what makes it weak, how the best researchers in the field set up their studies, and how I scored more than five thousand cases for evidence quality and for their fit with Scripture. The chapters ahead will bring you the cases and the data. This one gives you the measuring stick.

“Test everything; hold fast what is good.”
— 1 Thessalonians 5:21 (ESV)

That verse has been my working motto from the start. Paul did not tell the Thessalonians to believe everything. He did not tell them to reject everything. He told them to test. Good evidence is what survives the test. Bad evidence is what does not. And the honest researcher is the one who keeps testing, even when the results are not what he hoped for.²

What Makes an Account Strong

So what separates a strong NDE account from a weak one? Three things, working together.

The first is what researchers call a confirmed perception (or veridical perception, from the Latin word for “true”). This is the heart of the matter. A veridical perception is something the person saw, heard, or learned during their experience that was later confirmed to be accurate by someone else. The patient who described the surgical saw is a confirmed perception. The patient who felt deep peace is not, because peace is an inner state that no one else can check.³

The second marker of a strong account is independent corroboration. It is not enough for the patient to say, “I saw the nurse put my dentures in a drawer.” The nurse also has to confirm it. When the patient’s report matches what someone else saw, heard, or recorded, and when that someone else had no reason to go along with the story, we have something solid. The Bible itself lays down this standard: “A single witness shall not suffice against a person for any crime or for any wrong. . . . Only on the evidence of two witnesses or of three witnesses shall a charge be established” (Deuteronomy 19:15, ESV). The principle is simple. One voice can be wrong. Two voices, telling the same story with no way to have coordinated, are much harder to explain away.⁴

The third marker is timing. When did the perception happen? Was the person’s brain working normally, or was it shut down? If a patient describes events that took place while they were awake or lightly sedated, we cannot rule out ordinary hearing or brief consciousness. But if the perception happened during a period of documented cardiac arrest, when the brain had no measurable electrical activity, then we have a very different situation. The brain should not have been able to form memories at all, let alone accurate ones.⁵

When all three of these markers come together in one case, you have the kind of evidence that makes even a careful skeptic pause. It is one thing to say that a dying brain can produce vivid feelings. It is quite another to say that a dying brain can accurately observe a specific surgical instrument, describe a conversation happening in another room, or identify the drawer where a nurse placed a set of dentures.⁶

Looking Forward and Looking Back

Not all studies are designed the same way. The difference between a good study and a weak one often comes down to one question: did the researchers set up the test before the cases happened, or did they gather stories afterward?

A retrospective study looks backward. The researcher finds people who had NDEs months or years ago, interviews them, and tries to piece together what happened. This is how Raymond Moody began. He gathered stories from people who had nearly died, noticed the patterns, and published Life After Life in 1975.⁷ That book changed the conversation. But Moody himself admitted that his methods were informal. He did not use medical records in a systematic way. He did not compare his NDE patients to a control group of people who also nearly died but did not report an experience. His work opened the door, but it left many questions standing.

A prospective study looks forward. The researchers decide in advance exactly what they will measure, how they will collect data, and what will count as evidence. They set up in hospitals. They interview every surviving cardiac arrest patient, not just the ones who volunteer a story. They record the medical data before, during, and after the event. And in some cases, they place hidden images or sounds in the room so that if a patient claims to have floated above their body, the researchers can test whether the patient saw something they could not have known about otherwise.⁸

Prospective studies are the gold standard. They are harder to do, they take years, and they produce far fewer cases. But the cases they do produce are much stronger, because the data was collected carefully and in real time.

Cardiologist Michael Sabom did some of the earliest prospective work. He was skeptical of NDEs when he started. He wanted to test whether patients who claimed to have watched their own resuscitation were just guessing based on what they had seen on television. So he set up a simple comparison. He took a group of cardiac arrest survivors who reported watching their resuscitation from outside their bodies, and he took a control group of heart patients who had not had an NDE. He asked both groups to describe what happens during a resuscitation.⁹

The results were striking. Eighty percent of the control group made at least one major error. They confused the equipment. They described the wrong procedures. They guessed, and they guessed wrong. But among the NDE patients who claimed they had seen their resuscitation, not one made that kind of error. Some of them described details so specific to their own case that the description would not have matched any other resuscitation in the hospital.¹⁰

“When I asked him to tell me what exactly he saw, he described the resuscitation with such detail and accuracy that I could have later used the tape to teach physicians.” — Michael Sabom, M.D., Light and Death

Sabom’s work set a pattern that others would follow. Penny Sartori, an intensive care nurse in the United Kingdom, ran a five-year prospective study with similar results. NDE patients gave accurate descriptions. Non-NDE patients guessed wrong.¹¹

The largest prospective studies came later. In 2001, Dutch cardiologist Pim van Lommel and his team published a landmark study in The Lancet, one of the world’s most respected medical journals. They enrolled 344 consecutive cardiac arrest survivors across ten Dutch hospitals and interviewed each one within days. They found that 18 percent reported some memory from the period of unconsciousness. The study was carefully designed to rule out the most common objections: selection bias, delayed interviewing, and the influence of medications.¹²

Sam Parnia, a critical care physician, took the method even further. His first pilot study, published in 2001, enrolled sixty-three cardiac arrest survivors and found that 11 percent reported an NDE. He then led the much larger AWARE study from 2008 to 2012, covering fifteen hospitals in the United States and the United Kingdom. AWARE tracked 2,060 cardiac arrest events and used hidden visual targets placed on shelves near the ceiling, visible only from above, as a test of out-of-body claims.¹³

The AWARE study produced one verified case of a patient who accurately described events during a three-to-five-minute period of cardiac arrest, confirmed by the medical team. The patient described hearing a specific nurse by name and hearing staff say they had gotten him back. Parnia called this the first case in which a prospective study had timed and verified conscious awareness during a period with no heartbeat. Unfortunately, neither of the two out-of-body cases in AWARE happened in a room with a visual target, so the hidden images were never put to the test.¹⁴

Parnia went on to lead the even larger AWARE-II study, which followed 567 cardiac arrest patients and added real-time brain monitoring. The study found that one in five survivors had features of a recalled experience of death, and almost 40 percent had vague awareness without clear recall. The team also detected brain waves normally seen during conscious thought, including delta, theta, alpha, and beta waves, even in patients whose brains had flatlined. These spikes occurred up to an hour into resuscitation. For the first time, researchers had found brain-based markers that matched what millions of survivors had been describing for decades.³⁸

This is the reality of prospective NDE research. It is painstaking. It takes enormous patience. And the strongest cases remain rare, not because the experiences do not happen, but because so few cardiac arrest patients survive, and fewer still recall what happened.¹⁵

Bruce Greyson summarized three prospective studies with the same design. In the Dutch study, 18 percent of cardiac arrest survivors reported an NDE. In his own American study, the number was 15.5 percent. In the British study led by Parnia, it was 11 percent. Across all three, roughly 10 to 20 percent of people who survived cardiac arrest reported an experience of heightened consciousness during a period when, by every medical standard, their brains should not have been capable of producing any experience at all.¹⁶

Measuring What We Find

Once you start gathering NDE accounts, you need a way to sort them. Not all stories carry the same weight. Some include rich detail and outside confirmation. Others are sincere but vague. A good scoring system helps you see the difference without having to rely on gut feelings.

The Greyson NDE Scale

Bruce Greyson, a psychiatrist at the University of Virginia, built the first widely used measuring tool for NDEs in the early 1980s. He started with a list of the eighty features most often reported in the research. Then, through repeated testing with experiencers and other researchers, he narrowed it down to sixteen. Each feature is scored 0, 1, or 2, giving a total between 0 and 32. A score of 7 or higher counts as a near-death experience.¹⁷

The sixteen features cover four areas: changes in thinking (such as faster thoughts or a life review), changes in feeling (deep peace, a sense of being loved), unusual perceptions (seeing from outside the body, awareness of events elsewhere), and a sense of another world (meeting the dead, encountering a light, reaching a border). The scale does not measure how deeply a person was affected. Someone with a low score can still have a life-changing experience. What the scale does is give researchers a common language. It makes sure everyone is studying the same kind of experience.¹⁸

Twenty years after Greyson published the scale, two skeptical scholars from outside the NDE field asked to test it with a sophisticated statistical method. Greyson handed over his raw data from hundreds of experiencers. The analysis confirmed that the scale measured one consistent experience, the same for men and women, across ages and cultures. NDE Scale scores did not change depending on how many years had passed since the experience. The scale had passed a test designed by people who had no stake in making it look good.¹⁹

Evidence Quality: How I Scored 5,278 Cases

The Greyson Scale tells you whether an experience is a genuine NDE. But it does not tell you how strong the evidence is that the person’s mind was working apart from their brain. For that, I needed a different tool.

In my doctoral research, I developed an evidence-quality scale. I analyzed 5,278 near-death experiences, 832 from scholarly sources (published in academic journals and peer-reviewed books) and 4,446 from online databases. At the time, there was no standardized scoring system for veridical NDEs anywhere in the research literature. I had to build one from scratch.²⁰

The scale scored each case on three primary dimensions. First, the medical context: how severe was the crisis, and how well was it documented? A case involving documented cardiac arrest with a flat EEG scored far higher than a vague report of a close call. Second, the veridical quality: how specific were the observations, and how impossible would they have been to know through normal senses? A blind person reporting accurate visual details scored higher than a general description of a hospital room. Third, the corroboration strength: who confirmed the account, and when? Multiple medical professionals documenting their confirmation within twenty-four hours scored far higher than a family member recalling something years later.^20b

Because the two kinds of sources carry very different levels of documentation, I used two versions of the scale. The scholarly cases, which often came with full medical records reviewed by the researcher, could earn up to 90 points. The online self-reported cases, which usually lack that kind of documentation, could earn up to 50 points. But the three dimensions were the same in both. Every case was asked the same core questions.

The results were clear. About 30.7 percent of all cases (1,618 total) met the bar for “Exceptional” or “Strong” evidence. In 1,114 cases, the person reported accurate observations of events at a distance from their body. Thirty-three cases involved blind people who reported verified sight during their NDE.²¹

One finding surprised me the most. The worse the medical crisis, the better the evidence quality. The highest scores came from cases during documented cardiac arrest with EEG monitoring. A dying brain theory predicts the opposite. A brain that is shutting down should produce worse perception, not better. But the data said otherwise.²²

A related finding makes the point even sharper. Among 105 cases where the patient’s eyes were taped shut during surgery, 71.4 percent still produced high-quality veridical observations. These patients could not have been peeking. Their eyes were sealed. And yet they described what was happening in the room with accuracy that was later confirmed. If consciousness were merely a product of normal sensory input, this should not happen at all.^22b

I also found a large quality gap between the two kinds of sources. Scholarly cases, gathered by trained researchers who interviewed patients and checked records, scored far higher than self-reported online cases (31.2 versus 20.7 on my scale). That gap matters. It tells us that the strongest evidence comes from the most carefully gathered accounts, not from anonymous online forms. It also tells us that when we focus on the best-studied cases, the picture becomes more impressive, not less.²³

Testing for Biblical Fit

Evidence quality is only half the picture. As a Christian researcher, I also wanted to know how well each case lined up with what the Bible teaches. So I scored each case for biblical compatibility as well, looking at whether the content of the experience matched, contradicted, or was neutral toward the teachings of Scripture.

I also used a second tool for this part of the work: a nine-point framework developed by Shaun Tabatt (with Randy Kay), which maps the core elements that tend to appear in a genuine NDE. These include the death event itself, leaving the body, heightened senses, encountering other beings, a life review, reaching otherworldly realms, receiving special knowledge, returning to the body, and lasting changes afterward. The higher the score, the more likely the account reflects a genuine NDE rather than a dream or a vague impression. In my data, those who encountered Jesus specifically scored higher on Tabatt’s scale than the average, which makes intuitive sense: the more fully formed the experience, the more likely it includes a recognizable divine encounter.^24b

The results of the biblical-compatibility analysis were encouraging but also sobering. About 87 percent of the scholarly cases showed high or moderate compatibility with a biblical framework. Overall, about 71 percent of all cases lined up well with Scripture. But at least 1 in 10 carried clearly unbiblical content and had to be rejected on that basis. About 12 percent promoted some form of reincarnation, and about 19.6 percent promoted psychics or mediums.²⁴

These numbers tell us two things. The first is good news: the great majority of carefully studied NDEs are broadly compatible with what the Bible teaches about God, about the afterlife, about the reality of the soul. The second is a warning: not every story can be trusted. Some carry ideas that Scripture plainly rejects. This is why we need both an evidence score and a discernment score. The chapters on discernment and the field guide later in this book will give you the full framework for testing the spiritual content. For now, just keep this in mind: a strong evidence score does not by itself mean the experience came from God. More on that below.

Where Error Comes From

If we are serious about testing, we need to be honest about what can go wrong. Good researchers do not pretend that every account is reliable. They name the sources of error and then explain how careful method guards against them.

False Memory and Confabulation

The human brain fills in gaps. When memory is incomplete, the mind will sometimes stitch together a story from fragments, expectations, and things learned after the fact. This is called confabulation. The person is not lying. They truly believe their memory is accurate. But the details may have been added later, without the person realizing it. Anyone who has studied eyewitness testimony in courtrooms knows how common this is, even in people with healthy brains.²⁵

NDE researchers take this seriously. Several factors make NDE memories potentially vulnerable: cardiac arrest often causes amnesia, trauma can distort memory, and strong emotions can color what people recall. If we are honest, we have to admit that some NDE reports may contain memories that were shaped after the fact, filled in with details the person picked up later from conversations, television, or books.²⁶

But here is where the data pushes back. If NDE memories were confabulated, we would expect them to be vague, confused, and full of errors. Instead, they are typically the opposite. NDE patients report experiences that are sharper and more vivid than ordinary memory. Jeffrey Long found that 96 percent of experiencers rated their NDE as “definitely real.” Many say it was more real than anything they had ever experienced while awake. That is not the profile of a confused or fabricated memory. That is the profile of something the person lived through.³⁹

There is also a timing argument that works strongly in favor of NDE reports. In my research, I tracked when each account was first documented. In fourteen cases, the experiencer described what they saw or heard before anyone else told them what had happened. They wrote it down or told someone before any outside information could reach them. Their accounts were later confirmed to be accurate. An additional 1,250 cases were documented within twenty-four hours of the event, while the memory was fresh and before outside influences could reshape it. If these memories were built from overheard conversations or later suggestions, the timing does not fit.^25b

The Urge to Tell a Good Story

People want to be heard. When someone has a powerful experience, there is a natural pull to make the story a little bigger, a little more dramatic, with each telling. Researchers call this embellishment. It is not the same as lying. It is the slow drift that happens when a story gets told many times to many audiences.

Sabom guarded against this by seeking out what he called “virgin cases,” people who had not told their story to a mass audience. He avoided the support groups and the conference circuit. He wanted fresh accounts from people who had no reason to perform.²⁸ Van Lommel guarded against it by re-interviewing his patients at two years and again at eight years after their cardiac arrest. The stories did not change. This helped establish that NDE memories remain fixed, not growing with time.²⁹

Believer’s Bias

Some critics charge that NDE researchers are biased. They say the field attracts people who already believe in an afterlife, and that these researchers see what they want to see. This is worth taking seriously. Every scientist carries assumptions. The question is whether the method is strong enough to keep those assumptions from warping the results.³⁰

I think the charge is often unfair. Sabom started as a skeptic and became convinced only after his data forced him to change his mind. Parnia has said openly that when he began studying NDEs, he thought they were probably brain-based. Greyson has spent decades submitting his work to peer-reviewed journals and inviting independent review of his tools. These are not the habits of people who are afraid to be proven wrong.³¹

But I also admit that bias can run in both directions. A researcher who dismisses every NDE as a hallucination is just as biased as one who accepts every story as proof of heaven. The honest path is to let the data speak, to publish it openly, and to welcome the scrutiny of people who disagree. In my own work, I tried to guard against bias by scoring every case the same way, using fixed criteria set before I started. I scored online self-reports the same way I scored carefully gathered scholarly cases. The method was the same; the results were different. And I reported both.

There is another guard against bias that does not get enough attention: the experiences themselves resist expectations. Researchers have found that NDEs often contradict the beliefs the person held going in. Van Lommel noted that most of his Dutch patients had no expectation of dying, let alone of having a spiritual experience. Many did not believe in an afterlife at all. Parnia has pointed out that people from vastly different backgrounds, whether from rural Iran, London, Mexico, or the American Midwest, whether young or old, religious or atheist, describe the same core features in strikingly similar terms. If these experiences were produced by expectation, we would expect far more variety. Instead, the pattern holds across cultures and decades.⁴⁰

The Hardest Objections

Before we move on, let me put the strongest objections on the table and answer them directly.

“Anecdotes Are Not Data”

This is the objection I hear most often from scientists, and it has real force. A single story, no matter how moving, does not prove a general claim. I agree. That is why I did not build my research on single stories. I scored 5,278 cases using a consistent method. Other researchers have built databases of hundreds and even thousands. Jeffrey Long’s Near-Death Experience Research Foundation has gathered and analyzed more than 4,000 accounts using standardized questionnaires with built-in checks for consistency.³³ Janice Holden reviewed every published scholarly case of veridical out-of-body perception and found 107, from thirty-nine publications by thirty-seven different research teams. Of those cases, 92 percent were completely accurate. Only 8 percent contained any error at all.³⁴

When hundreds of anecdotes are gathered carefully, checked against records, and compared to control groups, they stop being mere anecdotes. They become a body of evidence. And a body of evidence can be weighed.

In my own data, 345 cases were verified by medical staff testimony, and those cases showed a 64.9 percent rate of high-quality evidence. Another 113 were verified by medical records alone, at a 68.1 percent high-quality rate. When multiple witnesses confirmed the same account, the rate climbed to 65.5 percent. Even a small number of cases had video or audio recording as backup. These are not campfire stories. These are documented reports, checked by the people who were in the room.^33b

“You Cannot Rerun the Experiment”

True. No one can schedule a cardiac arrest. Researchers cannot control when, where, or to whom an NDE will happen. This sets NDE research apart from most laboratory science, where you can repeat an experiment as many times as you like.

But this objection proves too much. Astronomy cannot rerun a supernova. Geology cannot replay an earthquake. History cannot repeat a battle. We do not throw out these fields just because their events are not repeatable on demand. What we do is gather the best evidence we can, compare cases, look for patterns, and see whether the explanations hold up. That is exactly what NDE researchers are doing.³⁵

“NDE Researchers Cherry-Pick the Good Cases”

Some critics suspect that researchers quietly set aside the cases that do not fit and publish only the impressive ones. This is a version of what scientists call the file-drawer effect.

Holden’s review addressed this directly. Of the 107 cases she examined, seven (about 6 percent) involved either minor or major error. She published those too. The fact that errors were reported and counted puts the file-drawer worry to rest. Researchers are not hiding the misses. When the errors are included and the accurate cases are still at 92 percent, the pattern is not the result of cherry-picking.³⁶

Prospective studies also guard against selection bias. Van Lommel enrolled every cardiac arrest survivor at his hospitals, not just the ones who came forward with a story. Parnia did the same. When you interview every patient, you cannot cherry-pick.³⁷

How to Read the Rest of This Book

You now have the tools. From here on, whenever I present an NDE case, I will tell you what kind of evidence it offers. I will note whether the perception was confirmed, who confirmed it, and when it happened relative to the medical crisis. When a case is strong, I will say so plainly. When it is weak, or when it relies on inner experience alone, I will say that too.

When I cite my own research, I will use the word “strong” or “exceptional” to mean that the case met my evidence-quality criteria: a specific perception, independently confirmed, during documented unconsciousness. I will also note where cases raise questions about their biblical fit, and I will point you to the discernment chapters for the fuller framework.

The next chapter, Chapter 8, looks at the out-of-body experience in depth. Chapter 9 brings the famous cases and the numbers. The data will speak for itself. But you will be able to hear it clearly because you now know what to listen for.

Let me close with a word about why this matters so much.

We are asking a question that touches the deepest hope of every human being: does the soul survive death? If the answer is yes, then everything changes. Death is not the end. The people we love who have died before us are not gone. And the God who made us body and soul keeps His promise to meet us on the other side.

That is a hope worth holding. But it is also a hope worth testing. I do not want to rest the weight of that promise on a story that cannot bear it. I would rather build on evidence that has been checked, challenged, and found to stand. I would rather have thirty solid cases than three thousand shaky ones.

In my own research, I found that the strongest cases do stand. They stand against the skeptics who say it is all brain chemistry. They stand against the critics who say it is all wishful thinking. And they line up, far more often than not, with what Scripture already teaches about the soul, the afterlife, and the God who made us for more than this world.

But we are not there yet. We are still sharpening our tools. The evidence chapters are ahead. For now, take the measuring stick. Hold it in your hand. Ask hard questions. Demand real answers. And as the cases come, weigh them carefully.

That is what the Bereans did. They did not reject what Paul told them. They did not accept it blindly, either. They examined the evidence. They checked it against what they already knew to be true. And in the end, many of them believed, not because they had stopped thinking, but because the evidence held up.

“Now the Berean Jews were of more noble character than those in Thessalonica, for they received the message with great eagerness and examined the Scriptures every day to see if what Paul said was true.”
— Acts 17:11 (NIV)

That is the spirit I want this book to breathe. Eagerness and examination. Hope and honesty. The chapters ahead will give you both.