5D Library › Power-Law Weighting and Bottom-Up Controllability in Multi-Tier Conditional Excitation Systems · 中文

Chapter 8 Practice and a Pre-Registered Trial

Note

This chapter is not career, investment or business advice. The checklist is a tool for organizing one’s thinking and is demonstrated on an invented example; it must not be used to rate, screen or “grade” anyone, and it does not replace career counselling, psychological help or legal advice.

Part 1 A four-step checklist

The models of Chapters 1–6 boil down to four questions. Their use is to break the vague question “where should I put my effort?” into steps that can be answered one at a time and corrected by facts.

  1. Find the weakest tier. In a multiplicative structure the weakest tier has the largest marginal value (Chapter 1). Which tier is weakest, and on what basis: observable facts (the opportunities in one’s industry and city, the mentors and peers one can reach) or self-assessment?
  2. Estimate the controllable inputs. Which inputs are actually available (courses, a mentor network, support for moving or commuting …), and which tier does each act on directly? Which tiers have no available input at all? By the author’s own assumption, C1C_1 is uncontrollable; C2C_2 often has no direct input either (Chapter 5).
  3. Assess the coupling. Besides the tier it acts on, which tiers does each input affect through the couplings? Could its effect on the overall outcome change direction under different coupling assumptions (Chapter 6)? Prefer inputs whose effect keeps its sign under every reasonable assumption.
  4. Design feedback. Review the state of each tier regularly (say monthly) and put the next round of investment into the tier that is then weakest and has an available input; record each round’s investment and result, and use the facts to correct the estimates of the coupling.

Calculation

The checklist applied to an invented person, “L.”. The five tier scores are made up and treated as the factors of the multiplicative model of Chapter 1; step 3 borrows the linear couplings of Chapter 6 (UP and CROWD) to compute the steady-state change in ln⁡S\ln S; step 4 compares, over 12 months, “always take the course” with “each month invest in the weakest reachable tier”, with one unit of investment per month and diminishing returns (p←p+0.06(1−p)p \leftarrow p + 0.06(1 – p)).

# The four-step check-list applied to a FICTIONAL person, "L." (Chapter 8). Layer scores
# (0 to 1, invented) are the factors of the multiplicative model of Chapter 1.
# 1 weakest layer: largest marginal gain P / p_i.  2 inputs available, and the layer
# each acts on.  3 coupling: does an input's steady-state effect on ln S keep its sign
# under the UP and CROWD couplings of p13?  4 feedback: 12 monthly reviews, one unit of
# effort a month, diminishing returns p <- p + 0.06 (1 - p); "always the course" versus
# "the weakest reachable layer each month".
import math

LAYERS = ["C1 prior", "C2 timing", "C3 environment", "C4 reputation", "C5 knowledge"]
p0 = [0.70, 0.55, 0.30, 0.50, 0.80]
INPUTS = {"evening course": 4, "mentor network": 3, "relocation support": 2}

def ln_s_change(j, crowd):        # layers 2-5 only (C1 frozen); unit input on layer j
    A = [[-1.0 if i == k else 0.5 if i == k + 1 else 0.4 if k == i + 1 else 0.0
          for k in range(4)] for i in range(4)]
    if crowd: A[1][3] = -0.9                      # study time crowds out networking
    X = [row[:] + [-1.0 if i == j - 1 else 0.0] for i, row in enumerate(A)]
    for c in range(4):
        for i in range(4):
            if i != c: X[i] = [a - X[i][c] / X[c][c] * b for a, b in zip(X[i], X[c])]
    return sum(X[i][4] / X[i][i] for i in range(4))

if __name__ == "__main__":
    P = math.prod(p0)
    print(f"Step 1  P = {P:.4f}; marginal gain P / p_i by layer:")
    for name, pi in zip(LAYERS, p0): print(f"        {name:15s} {pi:.2f}  {P / pi:.4f}")
    print("Step 2  inputs and the layer each acts on directly:")
    for k, j in INPUTS.items(): print(f"        {k:19s} -> {LAYERS[j]}")
    print("        nothing available acts on C1 or C2 directly")
    print("Step 3  steady-state change of ln S per unit input")
    print("        input                     UP    CROWD")
    for k, j in INPUTS.items():
        up, cr = ln_s_change(j, False), ln_s_change(j, True)
        print(f"        {k:19s} {up:+8.2f} {cr:+8.2f}"
              + ("   sign flips" if up * cr < 0 else ""))
    for policy in ("always the course", "weakest reachable"):
        p = p0[:]
        for month in range(12):
            best = max((2, 3, 4), key=lambda i: math.prod(p) / p[i])
            j = 4 if policy == "always the course" else best
            p[j] += 0.06 * (1 - p[j])
        print(f"Step 4  {policy}: P after 12 months {math.prod(p):.4f}"
              f" (x{math.prod(p) / P:.2f})")
        print("        layers " + " ".join(f"{x:.2f}" for x in p))
# Step 1  P = 0.0462; marginal gain P / p_i by layer:
#         C1 prior        0.70  0.0660
#         C2 timing       0.55  0.0840
#         C3 environment  0.30  0.1540
#         C4 reputation   0.50  0.0924
#         C5 knowledge    0.80  0.0577
# Step 2  inputs and the layer each acts on directly:
#         evening course      -> C5 knowledge
#         mentor network      -> C4 reputation
#         relocation support  -> C3 environment
#         nothing available acts on C1 or C2 directly
# Step 3  steady-state change of ln S per unit input
#         input                     UP    CROWD
#         evening course         +2.60    -0.85   sign flips
#         mentor network         +4.00    +1.70
#         relocation support     +4.25    +2.81
# Step 4  always the course: P after 12 months 0.0523 (x1.13)
#         layers 0.70 0.55 0.30 0.50 0.90
# Step 4  weakest reachable: P after 12 months 0.1079 (x2.33)
#         layers 0.70 0.55 0.60 0.58 0.80

In this invented example L.’s weakest tier is environment (0.30), whose marginal value of 0.154 is far above the others’. The available inputs are an evening course (acting on C5C_5), a mentor network (C4C_4) and relocation support (C3C_3); nothing acts on C1C_1 or C2C_2. Step 3 shows that the evening course changes ln⁡S\ln S by +2.60 under UP and by −0.85 under crowding-out, so its direction can flip; the mentor network and relocation support are positive under both. In step 4, “always take the course” raises the overall outcome only 1.13-fold in 12 months (C5C_5 from 0.80 to 0.90), while “invest in the weakest reachable tier each month” raises it 2.33-fold (C3C_3 from 0.30 to 0.60, C4C_4 from 0.50 to 0.58).

This is merely a consequence of the model, and all the numbers are invented. What it illustrates is a testable claim: in a multiplicative structure, aiming investment at the weakest reachable tier works better than concentrating it on tiers that are already strong. The trial below is designed to test that claim.

Part 2 A pre-registered randomized controlled trial

Question

For adults early in their careers who are having difficulty finding work, does structured support aimed at their “weakest reachable tier” (a mentor network, relocation or commuting support, or a targeted course, allocated by a pre-specified diagnostic) raise labour earnings in months 7–18 after randomization, compared with the same number of hours of general employability training?

Evidence

Why the effect of interest should be small. In a multidisciplinary meta-analysis, Eby and colleagues found that mentored people fared slightly better than non-mentored people across behavioural, attitudinal, health-related, relational, motivational and career outcomes, but the effect sizes were generally small (Eby et al. 2008). Summarizing more than 200 evaluations of active labour market programmes, Card, Kluve and Weber found average effects close to zero in the short run that became more positive two to three years after the programme; programmes emphasizing human capital accumulation had larger average gains, and effects were larger for women and the long-term unemployed (Card et al. 2018).

Design

Element Specification
Participants Aged 21–35, unemployed or clearly under-employed for at least 3 months, taking part voluntarily; recruited through public employment services. Not recruited from the author’s community, students or clients, nor from the clients or signed talent of Asia Influencer X or Yafu Global
Exclusions Currently in another active labour market programme; unable to give informed consent; not consenting to linkage with administrative earnings records
Diagnostic Before randomization, all participants complete a pre-registered five-tier questionnaire and objective indicators (for example the vacancy rate in their industry and area, the number of professional contacts they can reach); a rule written as code in advance picks the “weakest reachable tier” (chosen only among C3C_3, C4C_4, C5C_5; by the author’s model C1C_1 and C2C_2 have no direct input). No genetic data are collected
Randomization Individual, 1:1, stratified by site and by diagnosed weakest tier; central web-based allocation with concealment
Arm A 40 hours of structured support over 12 weeks aimed at the diagnosed weakest tier: C4C_4 → a network of mentors and industry contacts; C3C_3 → relocation or commuting support with job matching at the destination; C5C_5 → a targeted course
Arm B (active control) 40 hours of general employability training over 12 weeks (CVs, interviews, workplace communication), delivered by the same providers
Blinding Participants and providers cannot be blinded; the primary outcome comes from administrative records; analysts are blind to allocation until unblinding
Primary outcome Total labour earnings in months 7–18 after randomization (administrative tax or social insurance records, winsorized at the 99th percentile), in units of the control arm’s standard deviation
Secondary outcomes Working 30 or more hours a week at month 18; job quality; change in five-tier questionnaire scores; earnings in months 19–36 (long-term follow-up)
Analysis Analysis of covariance with arm as the factor and baseline earnings (12 months before randomization), site and diagnosed tier as covariates; intention to treat; two-sided α=0.05\alpha = 0.05

Sample size

Calculation

# Sample size for the Chapter 8 trial. Individual randomisation, 1:1.
# Arm A: 40 hours of structured support aimed at the participant's weakest REACHABLE layer
# (mentor network, relocation/commuting support, or a targeted course), assigned by a
# pre-specified diagnostic. Arm B: 40 hours of general employability-skills training.
# Primary outcome: total earnings in months 7-18 after randomisation from administrative
# records (winsorised at the 99th percentile), in SD units of the control arm.
# ANCOVA on baseline earnings (previous 12 months) with correlation rho shrinks the
# residual variance to (1 - rho^2). Normal approximation; two-sided alpha 0.05.
import math
from statistics import NormalDist
N = NormalDist()

def per_arm(d, rho, alpha=0.05, power=0.90):
    z = N.inv_cdf(1 - alpha / 2) + N.inv_cdf(power)
    return math.ceil(2 * z ** 2 * (1 - rho ** 2) / d ** 2) + 1

if __name__ == "__main__":
    print("analysable participants per arm, 90 % power")
    print("  effect (SD) | baseline correlation rho =  0.3    0.4    0.5")
    for d in (0.10, 0.15, 0.20, 0.25):
        print(f"     {d:.2f}     |" + " " * 27
              + "  ".join(f"{per_arm(d, r):5d}" for r in (0.3, 0.4, 0.5)))
    n = per_arm(0.15, 0.4); recruit = math.ceil(n / 0.90)
    print(f"chosen: 0.15 SD, rho = 0.4 -> {n} per arm")
    print("  allowing 10 % loss (withdrawn consent, failed record linkage):")
    print(f"  recruit {recruit} per arm, {2 * recruit} in all")
    print(f"  power at the same n if the true effect is 0.10 SD:"
          f" {N.cdf(0.10 / math.sqrt(2 * (1 - 0.4 ** 2) / n) - N.inv_cdf(0.975)):.0%}")
# analysable participants per arm, 90 % power
#   effect (SD) | baseline correlation rho =  0.3    0.4    0.5
#      0.10     |                            1914   1767   1578
#      0.15     |                             851    786    702
#      0.20     |                             480    443    396
#      0.25     |                             307    284    254
# chosen: 0.15 SD, rho = 0.4 -> 786 per arm
#   allowing 10 % loss (withdrawn consent, failed record linkage):
#   recruit 874 per arm, 1748 in all
#   power at the same n if the true effect is 0.10 SD: 58%

With a smallest effect of interest of 0.15 standard deviations, a correlation of 0.4 between baseline earnings and the outcome, and 90 % power, each arm needs 786 analysable participants; allowing about 10 % loss (withdrawn consent, failed record linkage), each arm recruits 874, 1,748 in all. If the true effect were only 0.10 SD, the same sample would have only about 58 % power. The smallest effect of interest of 0.15 is the author’s assumption, based on the meta-analyses above: labour market programmes usually have modest effects, and a larger value would make the trial insensitive to effects that are realistically possible.

What “90 % power” means. As in Volume 33, 90 % is the probability of reaching some verdict of benefit (arm A raises earnings) when the true effect exactly equals the smallest effect of interest (0.15 SD); in that case the decision rule reaches its strongest verdict, “at least 0.15 SD higher”, only about 2.5 % of the time. The decision program below confirms this by simulation.

Decision rule

Calculation

# Decision rule for the Chapter 8 trial, fixed before any data are collected.
# d = arm A minus arm B in the primary outcome (earnings in months 7-18, SD units of the
# control arm), from the pre-specified ANCOVA (baseline earnings, site, diagnosed layer);
# 95 % CI. POSITIVE d = higher earnings in arm A. Smallest effect of interest 0.15 SD.
# Gates: fewer than 80 % of arm-A participants received at least 30 of the 40 hours aimed
# at their assigned layer, or administrative records linked for fewer than 85 % of all
# randomised participants -> NOT INTERPRETABLE; serious harms judged related -> PAUSE.
import math, random
from statistics import NormalDist
SESOI = 0.15

def verdict(d, lo, hi, fidelity=0.92, linkage=0.95, harms=0):
    if harms:
        return f"d = {d:+.2f}: PAUSE for safety review and report - {harms} related harms"
    if fidelity < 0.80:
        return f"d = {d:+.2f}: NOT INTERPRETABLE - delivered as assigned: {fidelity:.0%}"
    if linkage < 0.85:
        return f"d = {d:+.2f}: NOT INTERPRETABLE - records linked for {linkage:.0%}"
    if lo >= SESOI:   v = "arm A raises earnings by at least 0.15 SD"
    elif lo > 0:      v = "arm A raises earnings; 0.15 SD not shown"
    elif hi < 0:      v = "arm A LOWERS earnings"
    elif lo >= -SESOI and hi <= SESOI: v = "no difference worth acting on"
    else:             v = "unclear: the interval is too wide to decide"
    return f"d = {d:+.2f} ({lo:+.2f} to {hi:+.2f}): {v}"

def label(lo, hi): return verdict(0, lo, hi).split(": ", 1)[1]

if __name__ == "__main__":
    for case in ((0.31, 0.21, 0.41), (0.18, 0.08, 0.28), (0.04, -0.06, 0.14),
                 (-0.14, -0.24, -0.04), (0.10, -0.06, 0.26), (0.25, 0.15, 0.35, 0.71),
                 (0.22, 0.12, 0.32, 0.90, 0.78), (0.20, 0.10, 0.30, 0.92, 0.95, 3)):
        print(verdict(*case))
    # How often does each verdict occur? Residual SD sqrt(1 - 0.4^2); summary statistics
    # are drawn from their exact sampling distributions (normal means, chi-square
    # variance).
    rng = random.Random(41); z = NormalDist().inv_cdf(0.975); sr = math.sqrt(1 - 0.4 ** 2)
    for n, true in ((786, 0.15), (786, 0.25), (786, 0.0)):
        tally = {}
        for _ in range(10000):
            d = rng.gauss(true, sr * math.sqrt(2 / n))
            df = 2 * n - 3
            s = sr * math.sqrt(rng.gammavariate(df / 2, 2) / df)
            se = s * math.sqrt(2 / n)
            v = label(d - z * se, d + z * se); tally[v] = tally.get(v, 0) + 1
        print(f"{n} per arm, true d = {true:.2f}; share of 10,000 simulated trials:")
        for k, c in sorted(tally.items(), key=lambda kv: -kv[1]):
            print(f"   {c / 100:5.1f} %  {k}")
# d = +0.31 (+0.21 to +0.41): arm A raises earnings by at least 0.15 SD
# d = +0.18 (+0.08 to +0.28): arm A raises earnings; 0.15 SD not shown
# d = +0.04 (-0.06 to +0.14): no difference worth acting on
# d = -0.14 (-0.24 to -0.04): arm A LOWERS earnings
# d = +0.10 (-0.06 to +0.26): unclear: the interval is too wide to decide
# d = +0.25: NOT INTERPRETABLE - delivered as assigned: 71%
# d = +0.22: NOT INTERPRETABLE - records linked for 78%
# d = +0.20: PAUSE for safety review and report - 3 related harms
# 786 per arm, true d = 0.15; share of 10,000 simulated trials:
#     86.8 %  arm A raises earnings; 0.15 SD not shown
#      7.8 %  unclear: the interval is too wide to decide
#      2.8 %  no difference worth acting on
#      2.6 %  arm A raises earnings by at least 0.15 SD
# 786 per arm, true d = 0.25; share of 10,000 simulated trials:
#     58.5 %  arm A raises earnings by at least 0.15 SD
#     41.4 %  arm A raises earnings; 0.15 SD not shown
#      0.0 %  unclear: the interval is too wide to decide
# 786 per arm, true d = 0.00; share of 10,000 simulated trials:
#     80.7 %  no difference worth acting on
#     14.4 %  unclear: the interval is too wide to decide
#      2.5 %  arm A raises earnings; 0.15 SD not shown
#      2.4 %  arm A LOWERS earnings

The example results are invented. If fewer than 80 % of arm A received at least 30 hours of support aimed at their assigned tier, or fewer than 85 % of participants could be linked to administrative records, the report says “not interpretable” whatever the results; if serious adverse events judged related to the intervention occur (for example serious financial hardship from relocating), the trial pauses for a safety review and reports them. Only if the lower limit of the effect’s confidence interval is at least 0.15 SD does the report say “arm A raised earnings by at least the smallest effect of interest”.

The simulation shows that with 786 per arm and a true effect of exactly 0.15, about 89 % of trials reach some verdict of benefit, but only 2.6 % the strongest; with a true effect of 0.25, 58.5 % reach the strongest verdict; with a true effect of zero, about 81 % conclude “no difference worth acting on”, and about 2.5 % each wrongly conclude benefit or harm.

What this trial cannot show

First, it cannot prove the five-tier model right. Even if arm A does better, that may be because its support is more personal or makes participants feel more valued, not because it targets the “weakest tier”; ruling that out needs a third arm that is personalized but not allocated by weakest tier, at the cost of half as many participants again. Second, arm A contains three very different interventions, and a positive result cannot say which one worked; subgroup analyses by diagnosed tier are only exploratory. Third, the validity of the diagnostic questionnaire is unknown: if it cannot identify the weakest tier accurately, a null result may reflect the diagnostic rather than the theory. Fourth, the trial does not touch C1C_1 or C2C_2 and collects no genetic data, so it says nothing about the “endowment tier”. Fifth, the primary outcome stops at month 18; the meta-analysis of Card and colleagues suggests that effects may appear later, and long-term follow-up is a secondary outcome. Sixth, the results apply only to similar populations and labour markets.

Ethics and conflicts of interest

The trial must be approved by an independent ethics committee, with informed consent (including linkage to administrative earnings records), and the protocol, diagnostic rule and analysis code must be registered before the first participant is enrolled (Nosek et al. 2018). Relocation support must be voluntary, and participants may decline to move at any time without losing other support; the team must provide an emergency fund and counselling for those who move. Data are de-identified before analysis, and diagnostic results are never given to employers or any third party. Anyone may withdraw at any time without giving a reason. The results will be published whether positive or negative.

Disclosure

The author uses the works of 5D Library, including this book, as the methodological basis of his community courses, the 5D Ascension Cards (Dimensional Shift Card Decks) and the Yafu AI knowledge base, and therefore has a commercial interest in the subject of this book. The author also uses this “bidirectional pyramid” framework in his community courses and consulting, and founded and runs Asia Influencer X and Yafu Global (discovery and management of influencers and talent), and therefore has a direct commercial interest in the subject of this book. The book promotes no course, card deck or service; it uses no data from the author’s students or clients; the Chapter 8 trial is designed to be run and analysed by researchers independent of the author’s business and does not recruit from the author’s community, students or clients (including the clients and signed talent of these businesses); all the code and the study protocol are published openly for anyone to check. All models in this book are offered solely for mathematical and statistical discussion; they constitute no career, educational, investment or talent-management advice, and the book makes no promise about the outcome of any person or project. To limit the influence of these interests, the diagnostic rule, intervention manuals and analysis code will be made public before enrolment, and the results, data and code will be published whatever the outcome.

References

Card, David, Jochen Kluve, and Andrea Weber. 2018. “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations.” Journal of the European Economic Association 16 (3): 894–931. https://doi.org/10.1093/jeea/jvx028.
Eby, Lillian T., Tammy D. Allen, Sarah C. Evans, Thomas Ng, and David L. DuBois. 2008. “Does Mentoring Matter? A Multidisciplinary Meta-Analysis Comparing Mentored and Non-Mentored Individuals.” Journal of Vocational Behavior 72 (2): 254–67. https://doi.org/10.1016/j.jvb.2007.04.005.
Nosek, Brian A., Charles R. Ebersole, Alexander C. DeHaven, and David T. Mellor. 2018. “The Preregistration Revolution.” Proceedings of the National Academy of Sciences 115 (11): 2600–2606. https://doi.org/10.1073/pnas.1708274114.

Fan, Y. (2026). Power-Law Weighting and Bottom-Up Controllability in Multi-Tier Conditional Excitation Systems: A Mathematical Synthesis of Endogenous Priors, Timing Windows, Network Topology and Reputational Games [多层条件激发系统中的幂律权重与逆向控制可及性] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23810093 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).