Spontaneous Cases Collection

How parapsychology researchers collect, document, and judge unplanned reports: vivid dreams, sudden impressions, apparitions, and meaningful coincidences that happen outside the lab.

Topic type: Field research method (case intake and archival practice)

Used for: spontaneous cases, crisis cases, apparitions, coincidences, and other one-off anomalies

Main risk: selection bias and hindsight reconstruction

Last updated: March 5, 2026

Read: core sources Related: Field Investigation Methods Related: Crisis Coincidence Reports

Overview

A “spontaneous case” is an unusual experience that happens outside a planned experiment. Think of a vivid dream that seems to predict a real event, a sudden feeling that something is wrong with a loved one, or a vision of someone who turns out to have just died. Researchers call these reports “spontaneous” because no one set them up. “Spontaneous cases collection” is the effort to gather these reports in a careful, organized way. The goal is to preserve the evidence and judge each case by clear standards, not just pass it along as a good story.¹

Quick take

Main product: a documented timeline with an evidence trail, not just a narrative.
Primary threats: hindsight, selection bias, and contamination through retelling.
Best upgrades: early records (notes, messages), independent corroboration, and standardized coding.
Bottom line: treat cases as data objects with a clear history, not as anecdotes.

Goals of systematic case collection

Why go to the trouble of collecting these reports in a structured way? There are four main reasons. First, structure preserves the features that make a case worth studying. Second, it keeps the raw experience separate from later interpretation. Third, it helps researchers estimate how biased the archive might be. Fourth, it turns individual stories into patterns that can be tested in the lab.

Why collect cases systematically?

Preserve evidential features: what was experienced, when it was recorded, and when verification happened.
Separate experience from interpretation: reduce “story drift” as the account gets retold.
Estimate bias space: track how cases enter the archive and, when possible, track the misses too.
Generate testable hypotheses: find recurring triggers, contexts, and informational limits that can be tested in a lab.

Limits (what case collection cannot guarantee)

No clean denominators by default: you usually don’t know how many non-matching experiences occurred and were never reported.
Weaker causal inference: uncontrolled settings leave many ordinary explanations open.
Meaning inflation risk: interpretation can grow beyond the original content unless early records exist.

Collection workflow (intake to documentation to corroboration)

Good case collection follows a clear sequence. You start by pinning down exactly what is being claimed. Then you document it with time-anchored records. Then you check whether any witnesses can confirm the story independently.

1) Intake: define the claim

Write a one-sentence claim that can be checked: “Person X experienced Y at time T; later, event E occurred at time U.”
Record the witness account word for word before discussing what it means.
Log the first disclosure: who was told, when, and exactly what details were shared.

2) Documentation: preserve time anchors

The most valuable records are ones created before the outcome was known. These include diaries, letters, time-stamped messages, call logs, work schedules, medical timelines, and receipts. Any third-party document that pins down when the experience was reported helps block hindsight reconstruction.²

3) Corroboration: independence first

Interview corroborators separately when possible.
Map contamination: figure out who learned which details from whom after the fact.
Document uncertainty explicitly. Unknowns are part of the case record, not something to hide.

4) Alternative explanations: stress test ordinary routes

Perceptual and sleep-related: hypnagogia (the half-asleep state), grief effects, misperception, seeing patterns in random noise.
Informational: rumor, subtle cues, prior probability, or ordinary inference mistaken for psi.
Statistical: coincidence, base-rate neglect, selective memory.
Social: suggestion, conformity, escalation through retelling.

Coding and classification

Coding means assigning each case a set of labels based on specific features. Without coding, a collection is just an anthology of stories. With coding, researchers can compare cases and look for patterns. For example: are the best-documented cases also the most specific? Do cases with independent witnesses cluster in certain contexts?¹

Why coding matters

Coding turns a pile of stories into analyzable data. Investigators can compare cases on timing precision, documentation quality, specificity, relationship closeness, and how plausible ordinary information routes were.¹

Practical coding dimensions

Timing: exact timestamp vs. same-day vs. vague.
Documentation tier: contemporaneous record; early witness account; late recollection only.
Specificity: checkable details vs. broad mood (“something felt wrong”).
Information control: how likely was it that the person could have known through ordinary means?
Independence: isolated witnesses vs. people who all heard the same retelling.

What makes a case strong (and what weakens it)

Not all cases carry the same weight. Some features push a case toward being credible evidence. Others undermine it. Knowing the difference is the core skill in evaluating this kind of research.

Strengtheners

Contemporaneous record: the account exists in a dated form before the outcome was known.
Independent corroboration: early witnesses can confirm when and what was disclosed.
Bounded opportunity: ordinary information pathways are implausible or can be ruled out.
Specificity: claims that could clearly be wrong, not just emotionally resonant ones.

Weakeners

The first account appears only after the outcome is widely known.
All corroboration comes from people exposed to the same retelling.
Key specifics “appear” later as the story is repeated.
Misses are not recorded, making it impossible to judge how often similar experiences occur without a match.

Skeptical critiques

Skeptics argue that compelling case stories can arise from entirely ordinary processes. People have many experiences every day. Some will coincidentally match later events just by chance. Memory then reshapes the story to fit the outcome. Social retelling reinforces it further. From this view, a collection of “hits” is just the visible tip of a much larger pool of unreported misses and vague impressions that never got written down.³

Core skeptical objections

Selection bias: archives fill up with memorable matches. The forgettable non-matches never arrive.
Hindsight: narratives sharpen and become more specific after the outcome is known.
Unmeasured denominators: without a baseline, there is no way to judge how often “matching” experiences happen by chance.
Contamination: retelling produces apparent corroboration that is not truly independent.

Evaluation and refutation

How do you judge whether a collection is credible? And what would it take to show that a case, or a whole collection, does not support a psi explanation? There are clear answers to both questions.

How to evaluate a collection’s credibility

Intake discipline: standardized questions, word-for-word capture, and explicit acknowledgment of uncertainty.
Provenance: a clear chain from the first account to the archive entry.
Documentation: time-stamped records prioritized over late recollection.
Independence: corroboration mapped for contamination pathways.
Bias tracking: any attempt to record misses or estimate how selective the reporting was.

What would count as strong refutation (practically)

At the level of a single case: if key content cannot be shown to predate verification, or if ordinary information routes remain plausible, the case is weak evidence for psi.

At the level of a whole collection: if systematic intake and coding show that the “hits” are no more specific or time-anchored than coincidence and bias would predict, then psi interpretations are weakened. The strongest collections try to measure how biased the reporting was, rather than assuming the bias away.

Sources & notes

Irwin, H. J., & Watt, C. A. An Introduction to Parapsychology (5th ed.). Used here for spontaneous-case context, evidential criteria, and methodological cautions. ↩
A History of Parapsychology (in Psychic Exploration: A Challenge for Science). Used here for historical context on organized case collection and archival practice in psychical research. ↩
Parapsychology (Beginner’s Guides). Used here for discussions of coincidence, selection bias, and the limits of anecdotal inference. ↩