外气行为分与生理指标的共变,可能是评分者的锚定效应吗

Written by

in

Language

Evidence level: C

外气研究里「行为评分越高、发功态生理指标变化越大」这种共变,未必证明外气真实存在——它可能只是评分者看到「表演得像发功」之后的刻板印象打出来的锚定伪影。

正文

外气研究中有一个常见的论证链条:先让评分者看录像,给治疗师的「发功行为」打分(比如外气强度量表的行为分),再把这些行为分和受试者同期的心率变异性(HRV)、脑电等生理指标做相关。如果共变显著,就宣称「行为分越高,生理反应越强,说明外气有真实的生理对应物」。这个链条看起来漂亮,但它有一个几乎没人排查过的漏洞:共变的真正驱动力,可能既不是外气也不是受试者的生理,而是评分者自己——他们看到录像里的人摆出熟悉的发功姿态,就会按「这是真发功」的刻板印象给高分,行为分实际上锚定在「表演像不像发功」上,而不是「外气有没有发生」上。

这不是空想。评分者偏倚在临床研究里有非常扎实的实证:2025 年一项纳入 66 项随机试验的元流行病学研究发现,非盲评分者平均把干预效应夸大约 29%,主观结局上偏差更大;2023 年对一项卒中试验的分析显示,评分者的组别猜测正确率显著超过随机水平,且猜错猜对与结局评分显著相关。更直接对应本题的是 2015 年的一项心理治疗试验:治疗师自己的评分始终比独立盲法评估者更乐观,差异方向与治疗师的期望完全一致——说明「当事人打分」会被期望拉偏。2000 年还有一项经典方法学研究,用患者自评与临床医师评分的效应量差来量化评分者偏倚,给出了可操作的检偏工具。外气领域自己的数据也埋着线索:2008 年的外气膝骨关节炎随机对照试验里,两位真治疗师的结果不一致,只有一位优于「模仿发功」的假治疗师对照组——这本身就是「治疗师个体差异可能部分来自表演风格差异」的暗示。生理端方面,2005 年的外气 HRV 随机对照研究和同期 HRV 非线性分析确实记录到了组间自主神经差异,但那些研究的评分环节并未解决锚定问题。

要把共变从「锚定伪影」里洗出来,需要一个新的检验设计:治疗师组 vs 演员模仿组,同协议录像 + 盲法评分 + 共变强度组间对比。具体做法:让真外气治疗师和受过发功姿态训练、但按设计没有发功意图的演员,在完全相同的协议下录像,受试者同样佩戴 HRV/EEG 设备;然后让对分组毫不知情的评分者统一打行为分。关键的裁决指标是行为分×生理指标的共变强度在两组之间的对比:如果共变只在治疗师组出现而演员组没有,说明共变对「真实发功」有某种特异性,值得进一步追;如果演员组同样出现——甚至共变强度差不多——那么这条共变最经济的解释就是评分者的刻板印象锚定,行为分测的是「像不像发功」,不是「是不是发功」。还可以加一个正向操纵:让演员「演得更像」,如果行为分随之升高而生理共变不变,锚定解释就基本坐实。

这个检验的价值在于它便宜且难以反驳:不需要否认外气,只是要求在宣称「行为-生理共变即外气证据」之前,先排除一个几十年来在临床评分研究里被反复证实的混淆源。对研究者来说,盲法评分者、用演员模仿做阴性对照,是比「两位治疗师结果不一致再研究研究」更根本的修法。

证据与局限

已验证: 非盲评分者会系统性夸大效应估计(2025 元流行病学,66 试验);评分者破盲与结局评分相关(2023 卒中试验);治疗师自评受期望拉偏(2015 IPT 试验)。

部分验证: 外气随机对照研究中存在假治疗师模仿设计先例(2008 膝骨关节炎试验),且真治疗师间结果不一致;外气期间 HRV 有组间差异的报告存在(2005),但评分环节未做锚定排查。

待验证: 「行为分×生理共变强度在治疗师组与演员模仿组间无差异(或演员组同样出现)」这一裁决性预测目前没有任何直接研究检验过;行为分对外气行为的评分者间信度本身也缺乏系统数据。

相关阅读

证据等级: C
参考文献: PMID 40258524, PMID 36169032, PMID 25782762, PMID 10788756, PMID 18654733, PMID 16048810, PMID 16173532

证据等级:A=有PubMed直接支撑,B=科学类比/间接证据,C=传统文化共识/初步探索

L1 网络状态:该主题尚未进入 L1 对应网络。经查 N200 研究内核 CORRESPONDENCE_NETWORK.md 与 EDGE_TABLE_science-yixue.md,当前不存在「外气行为分×生理指标共变×评分者锚定」对应边;EDGE_TABLE 中仅有的相邻边为 HRV 级 A 边(挂 abdominal-breathing-body-cavity-resonance,PMID 10593633)与太极/纳音级 B 边,均与本命题的评分者偏倚裁决无直接对应。已将缺边信息登记至 science-yixue-l1-gaps.md。


免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Convariance of External Qi Behavior Scores and Physiological Indicators: Could It Be Rater Anchoring Bias?

Evidence level: C

In Feng Shui, external Qi refers to environmental energy patterns perceived as influencing human states. The observed covariation—where higher behavioral scores correlate with greater physiological indicator changes—does not necessarily prove the existence of external Qi. It may instead represent an anchoring artifact: a cognitive bias where raters, perceiving a performance that “resembles” active Qi emission, unconsciously inflate their scoring due to established stereotypes.

Evidence and Limitations (证据与局限)

  • Evidence Level: C (structural analogy; not direct measurement of Qi itself).
  • Limitation: This explanation relies on the structural analogy between behavioral performance and physiological response, rather than direct measurement of Qi itself.

Body

A common argument chain in external Qi research proceeds as follows: raters watch video recordings and assign scores to the therapist’s “emission behavior” (e.g., a behavioral score on an External Qi Intensity Scale); these behavioral scores are then correlated with the subjects’ concurrent physiological measures such as heart rate variability (HRV) and electroencephalogram (EEG). If the covariation is significant, it is claimed that “higher behavioral scores correspond to stronger physiological responses, indicating that External Qi has genuine physiological correlates.” This chain appears elegant, yet it contains a nearly unexamined flaw: the true driver of the covariation may be neither External Qi nor the subjects’ physiology, but rather the raters themselves. When raters observe familiar emission postures in the video, they tend to assign high scores based on the stereotype “this is true emission,” so the behavioral score becomes anchored to “how well the performance resembles emission” rather than “whether emission actually occurred.”

This is not mere speculation. Rater bias in clinical research is well-documented: a 2025 meta-epidemiological study including 66 randomized trials found that non-blinded raters inflated intervention effects by an average of 29%, with greater bias on subjective outcomes; a 2023 analysis of a stroke trial showed that raters’ group-guessing accuracy significantly exceeded chance, and their guesses were significantly related to outcome ratings. More directly relevant is a 2015 psychotherapy trial: therapists’ own ratings were consistently more optimistic than those of independent blinded assessors, with the direction of difference matching the therapists’ expectations—demonstrating that “patient ratings” are pulled by expectation. A 2000 classic methodological study quantified rater bias using the effect-size difference between patient self-ratings and clinician ratings, providing an actionable tool for detecting bias. External Qi’s own data also contain clues: in the 2008 randomized controlled trial of external Qi for knee osteoarthritis, results from two real therapists were inconsistent, with only one outperforming the “imitating emission” sham-therapist control—this itself hints that “individual therapist differences may partly stem from performance style differences.” On the physiological side, the 2005 external Qi HRV randomized controlled trial and its contemporaneous nonlinear HRV analysis did record intergroup autonomic differences, yet those studies’ rating procedures failed to address the anchoring problem.

To extract the true covariation from this “anchoring artifact,” a new test design is required: therapist group vs. actor imitation group, using identical protocol video + blinded rating + comparison of covariation strength between groups. The procedure: record both real external Qi therapists and actors trained in emission postures but designed without emission intent, under the exact same protocol, with subjects wearing HRV/EEG devices; then have raters who are unaware of the grouping assign behavioral scores. The critical decision metric is the strength of covariation between behavioral score and physiological measures compared across the two groups: if covariation appears only in the therapist group while the actor group shows none, this suggests covariation is specific to “true emission” and warrants further pursuit; if the actor group also exhibits covariation—even with similar strength—then the most parsimonious explanation is that the covariation reflects the raters’ stereotypical anchoring, meaning the behavioral score measures “how emission-like” rather than “whether emission occurred.” A positive manipulation can further confirm this: have actors “perform more convincingly”; if behavioral scores rise while physiological covariation remains unchanged, the anchoring explanation is essentially confirmed.

The value of this test lies in its cost-effectiveness and resistance to refutation: it does not deny External Qi, but requires that before claiming “behavior–physiology covariation constitutes evidence of External Qi,” one first excludes a confounding source repeatedly confirmed over decades in clinical rating research. For researchers, using blinded raters and actor imitation as negative controls represents a more fundamental methodological improvement than simply “repeating the study with two therapists who gave inconsistent results.”

Evidence and Limitations

Verified: Non-blind raters systematically overestimate effect sizes (2025 meta-epidemiology, 66 trials); rater unmasking correlates with outcome scoring (2023 stroke trial); therapist self-rating is biased by expectation (2015 IPT trial).

Partially Verified: Precedents for sham therapist mimicry exist in external Qi randomised controlled trials (2008 knee osteoarthritis trial), and results differ between true therapists; reports of inter-group differences in HRV during external Qi (2005) exist, but the rating phase lacked anchoring checks.

Pending Verification: The adjudicative prediction that “the strength of the co-variation between behavioural scores and physiological measures does not differ between therapist groups and actor mimicry groups (or the same occurs in the actor group)” has not yet been directly tested by any study; furthermore, systematic data on inter-rater reliability of behavioural scores for external Qi behaviour is itself lacking.

Related Reading

Evidence level: C
References: PMID 40258524, PMID 36169032, PMID 25782762, PMID 10788756, PMID 18654733, PMID 16048810, PMID 16173532

Evidence level: A = direct PubMed support; B = scientific analogy / indirect evidence; C = traditional consensus / preliminary exploration

L1 network status: This topic has not yet entered the L1-corresponding network. Examination of the N200 research core CORRESPONDENCE_NETWORK.md and EDGE_TABLE_science-yixue.md reveals no edge corresponding to “external Qi behavior branch × physiological indicator covariation × rater anchor.” The only adjacent edges in the EDGE_TABLE are HRV-level A edges (linked to abdominal-breathing-body-cavity-resonance, PMID 10593633) and Taiji/Nayin-level B edges; neither directly corresponds to the rater bias adjudication of this proposition. The missing-edge information has been logged in science-yixue-l1-gaps.md.


Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *