5D Library › Embodied Cognition, Bernstein's Degrees-of-Freedom Problem and the Control Theory of Motor Synergies · 中文

Chapter 8 Practice and a Pre-Registered Trial

Note

The checklist in this chapter is a thinking tool for analysing data, not a prescription for training, rehabilitation or treatment. The two participants in the example and all the data are fictional, used only to demonstrate the calculations. The trial protocol is a design for independent researchers to review and carry out, not a promise that any training method works. Anyone with an injury, pain or a neurological condition should consult a qualified professional.

Part 1: a four-step embodied-synergy checklist

The content of the first six chapters can be condensed into four questions. They apply to any data in which several elements accomplish a lower-dimensional task: joints and a fingertip, fingers and a total force, muscles and a joint torque.

  1. Write down the task variable and the redundancy. Which variable is supposed to be stabilised (fingertip position, cursor, total force)? How many elements are there (nn), and how many dimensions has the task variable (dd)? At which moment is the analysis made? The task variable should be written down before looking at the data; otherwise it is easy to “find” a variable that happens to be stabilised (Chapter 2).
  2. Decompose variability per degree of freedom, with intervals. Compute VUCMV_{\mathrm{UCM}} and VORTV_{\mathrm{ORT}} per degree of freedom, ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}}) or ΔV\Delta V, and give bootstrap confidence intervals; with only about 20 trials the intervals can be wide (Chapter 2).
  3. Check coordinates and a null model. Does the conclusion survive another reasonable choice of coordinates (relative versus absolute angles, rescaling, different units for different joints)? Under a null model of independent elements with the same number of trials, what range does the index fall in (Chapter 2)? With EMG and NMF, also check which muscles were recorded and whether task constraints can explain the low dimensionality (Chapter 5).
  4. Return to the task outcome. Finally look at the error of the task variable itself and at performance under new conditions. A large synergy index does not mean the task is done well; in the model of Chapter 6 the synergy index changed greatly while the task error did not change.

Calculation

The checklist applied to fictional data from the task of this chapter: two joint angles drive a one-dimensional cursor (cursor =0.8q1+0.6q2= 0.8q_1 + 0.6q_2), with 24 trials from each of two participants. A varies greatly along the UCM (SD 4.0°) and by 1.6° in the ORT direction; for B the figures are 1.5° and 0.9°.

# The four-step embodied-synergy checklist applied to SIMULATED data from the Chapter 8
# task: two joint angles (deg) drive one cursor, y = 0.8 q1 + 0.6 q2, so n = 2, d = 1.
# Two invented participants, 24 trials each at one time point of the movement.
# Step 1: task variable and redundancy. Step 2: per-DOF variances, ln(Vucm/Vort) with a
# bootstrap CI. Step 3: coordinate check (joint 2 expressed in other units/scaling)
# and a null model. Step 4: back to the outcome (task error), not the index.
import math, random
J = (0.8, 0.6)

def decompose(trials, j=J):
    n = math.hypot(*j); e = (j[0] / n, j[1] / n); u = (-e[1], e[0])
    m = [sum(t[i] for t in trials) / len(trials) for i in (0, 1)]
    dev = [(t[0] - m[0], t[1] - m[1]) for t in trials]
    vo = sum((d[0] * e[0] + d[1] * e[1]) ** 2 for d in dev) / (len(dev) - 1)
    vu = sum((d[0] * u[0] + d[1] * u[1]) ** 2 for d in dev) / (len(dev) - 1)
    return vu, vo                              # both 1-D here, so already per DOF

def boot_ci(trials, rng, reps=2000):
    est = []
    for _ in range(reps):
        vu, vo = decompose([rng.choice(trials) for _ in trials])
        est.append(math.log(vu / vo))
    est.sort(); return est[int(0.025 * reps)], est[int(0.975 * reps) - 1]

def person(rng, s_ucm, s_ort, n=24):
    n0 = math.hypot(*J); e = (J[0] / n0, J[1] / n0); u = (-e[1], e[0])
    out = []
    for _ in range(n):
        a, b = rng.gauss(0, s_ucm), rng.gauss(0, s_ort)
        out.append((20 + a * u[0] + b * e[0], 30 + a * u[1] + b * e[1]))
    return out

if __name__ == "__main__":
    rng = random.Random(43)
    people = {"A": person(rng, 4.0, 1.6), "B": person(rng, 1.5, 0.9)}
    print("step 1: n = 2 joints, d = 1 task variable -> UCM and ORT are each 1-D")
    print("step 2: per-DOF variances (deg^2) and ln(Vucm/Vort), 95 % bootstrap CI")
    for k, t in people.items():
        vu, vo = decompose(t); lo, hi = boot_ci(t, rng)
        print(f"  person {k}: Vucm {vu:6.2f}  Vort {vo:5.2f}"
              f"  ln ratio {math.log(vu / vo):+.2f}  ({lo:+.2f} to {hi:+.2f})")
    print("step 3a: joint 2 recorded in other units (x 2) -> ln ratio changes:")
    for k, t in people.items():
        vu, vo = decompose([(a, 2 * b) for a, b in t], (J[0], J[1] / 2))
        print(f"  person {k}: ln ratio {math.log(vu / vo):+.2f}")
    null = []
    for _ in range(4000):                      # null: independent joints, equal SDs
        t = [(rng.gauss(0, 2), rng.gauss(0, 2)) for _ in range(24)]
        vu, vo = decompose(t); null.append(math.log(vu / vo))
    null.sort()
    print("step 3b: null model (24 trials, independent joints with equal SDs):")
    print(f"  95 % of ln ratios fall in {null[100]:+.2f} to {null[3899]:+.2f}")
    print("step 4: back to the outcome, SD of the cursor (task units):")
    for k, t in people.items():
        y = [J[0] * a + J[1] * b for a, b in t]; m = sum(y) / len(y)
        sd = math.sqrt(sum((v - m) ** 2 for v in y) / (len(y) - 1))
        print(f"  person {k}: cursor SD {sd:.2f}")
# step 1: n = 2 joints, d = 1 task variable -> UCM and ORT are each 1-D
# step 2: per-DOF variances (deg^2) and ln(Vucm/Vort), 95 % bootstrap CI
#   person A: Vucm  22.74  Vort  2.14  ln ratio +2.36  (+1.83 to +2.95)
#   person B: Vucm   2.32  Vort  1.21  ln ratio +0.65  (-0.34 to +1.72)
# step 3a: joint 2 recorded in other units (x 2) -> ln ratio changes:
#   person A: ln ratio +3.22
#   person B: ln ratio +1.58
# step 3b: null model (24 trials, independent joints with equal SDs):
#   95 % of ln ratios fall in -0.84 to +0.87
# step 4: back to the outcome, SD of the cursor (task units):
#   person A: cursor SD 1.46
#   person B: cursor SD 1.10

Step 1: n=2n = 2 and d=1d = 1, so the UCM and ORT are each one-dimensional. Step 2: A’s ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}}) is +2.36 (95 % bootstrap interval +1.83 to +2.95), B’s +0.65 (−0.34 to +1.72), and B’s interval includes zero. Step 3: if the second joint is recorded in other units (values multiplied by 2), A’s log ratio becomes +3.22 and B’s +1.58 — the same movements, a conclusion that changes with the unit; under the null model (24 trials, two independent joints with equal SDs) 95 % of log ratios fall between −0.84 and +0.87, and B’s estimate lies in that range. Step 4: the SD of the cursor is 1.46 for A and 1.10 for B. A, with the larger synergy index, actually does the task worse than B. This is why Part 2 makes task accuracy, not the synergy index, the primary outcome.

Part 2: a pre-registered randomised controlled trial

Question

In a redundant tracking task in which two joints drive a one-dimensional cursor, do “external task-outcome focus instructions + redundancy-manifold feedback”, compared with “internal fixed-joint-angle instructions”, improve tracking accuracy on a new trajectory 24 hours later (transfer)?

Evidence

Existing evidence. The meta-analysis of external focus reports g=0.584g = 0.584 for transfer tests (Chua et al. 2021), and the estimate corrected for reporting bias is below 0.1 (McKay et al. 2024) (Chapter 6). UCM studies show that practice changes synergy structure (Yang and Scholz 2005), and optimal feedback control theory predicts that paying only for the task produces variability along the UCM (Todorov and Jordan 2002). But the author found no pre-registered trial that randomly compares these two forms of instruction in a redundant tracking task with transfer accuracy as the single primary outcome.

Design

Element Specification
Design Two-arm parallel laboratory trial, individually randomised 1:1, stratified by pre-practice error (above / below the median) and handedness
Participants Adults aged 18–65, recruited through universities and public research-participant platforms; not recruited from the author’s community, students or customers
Exclusions Musculoskeletal injury or surgery of the upper limb within 6 months, known neurological disease, pain during the task (stop at once; counted as “missing”)
Task Seated, forearm on a low-friction support; horizontal shoulder flexion–extension and elbow flexion–extension recorded by motion capture; cursor height =0.8q1+0.6q2= 0.8q_1 + 0.6q_2 (converted to screen millimetres), tracking a target trajectory made of several sinusoids
Practice Day 1: 2 pre-practice tests (no augmented feedback), practice 10 blocks × 6 trials
Intervention arm External instruction “keep the cursor on the target”; after each trial, a display of the joint-angle points of that trial and the UCM line (“any combination on the line is equally good”)
Control arm Internal instruction “keep the two joints moving together in the prescribed ratio”; after each trial, a display of each joint’s deviation from the prescribed joint trajectory
Primary outcome RMS cursor-tracking error (mm) on a new trajectory without augmented feedback 24 h later — transfer accuracy
Secondary outcomes Error on the practised trajectory 24 h later (retention); ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}}) and ΔV\Delta V in the transfer test (pre-specified in relative joint angles, in degrees); joint ranges of motion; reports of pain and discomfort; self-reported attentional focus
Sample size 113 analysable and 133 recruited per arm, 266 in all
Analysis ANCOVA of the primary outcome on arm + pre-practice error + stratification variables, reporting the adjusted mean difference dd (control minus intervention; positive means the intervention arm is better) with its 95 % confidence interval, two-sided α=0.05\alpha = 0.05

Both arms see exactly the same real-time cursor and target during practice, with the same number of trials, duration and payment; they differ only in the instruction and the summary display after each trial. Both forms of instruction are designed so that “if followed, the cursor stays on the target”: the joint ratio prescribed for the control arm is itself a feasible solution. The training materials are written by independent researchers and use no content from the author’s courses. The staff who run the tests and compute the outcomes do not know the participants’ allocation; participants cannot be blinded to the instruction they receive, but are not told the study hypothesis.

Why only one primary outcome. The framework wanted to compare both transfer accuracy and ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}}). This book makes only the former primary: first, the purpose of training is to do the task well under new conditions, which is what transfer accuracy measures; second, the model of Chapter 6 shows the synergy index changing by more than six units while the task error does not change at all, and Chapter 2 and Part 1 of this chapter show that it depends on coordinates and units; third, two primary outcomes would need a correction for multiple comparisons, enlarging the sample and making the conclusion harder to interpret. The log ratio is reported as a secondary outcome to help explain mechanisms.

Sample size

Calculation

# Chapter 8 trial: sample size. Primary outcome: RMS cursor-tracking error (mm) on a
# NEW trajectory 24 h after practice, without augmented feedback (transfer), analysed
# by ANCOVA with the pre-practice RMS error as covariate. Assumptions (not estimates):
# SD 4.0 mm, baseline-outcome correlation 0.5, smallest effect of interest 1.5 mm,
# two-sided alpha 0.05, power 90 %, 15 % attrition. Part 2 checks power by simulation.
import math, random
from statistics import NormalDist
Z = NormalDist().inv_cdf

def n_ancova(sd, rho, delta, power=0.90, alpha=0.05):
    return math.ceil(2 * (Z(1 - alpha / 2) + Z(power)) ** 2 * sd ** 2 * (1 - rho ** 2)
                     / delta ** 2)

def ancova(y, x, g):
    """OLS of y on [1, g, x]; returns the group coefficient and its SE."""
    n = len(y); mx, mg, my = sum(x) / n, sum(g) / n, sum(y) / n
    sxx = sum((a - mx) ** 2 for a in x); sgg = sum((a - mg) ** 2 for a in g)
    sxg = sum((a - mx) * (b - mg) for a, b in zip(x, g))
    sxy = sum((a - mx) * (b - my) for a, b in zip(x, y))
    sgy = sum((a - mg) * (b - my) for a, b in zip(g, y))
    det = sxx * sgg - sxg ** 2
    bg = (sxx * sgy - sxg * sxy) / det; bx = (sgg * sxy - sxg * sgy) / det
    res = sum((b - my - bg * (c - mg) - bx * (a - mx)) ** 2 for a, b, c in zip(x, y, g))
    return bg, math.sqrt(res / (n - 3) * sxx / det)

def trial(n, delta, rng, sd=4.0, rho=0.5):
    x, y, g = [], [], []
    for arm in (0, 1):                    # arm 1 = intervention, lower error by delta
        for _ in range(n):
            b = rng.gauss(0, 1); e = rng.gauss(0, 1)
            x.append(12 + sd * b)
            y.append(10 + sd * (rho * b + math.sqrt(1 - rho ** 2) * e) - delta * arm)
            g.append(arm)
    bg, se = ancova(y, x, g)
    return -bg, se                        # positive = intervention better

if __name__ == "__main__":
    n = n_ancova(4.0, 0.5, 1.5); rec = math.ceil(n / 0.85)
    print(f"main plan: {n} analysable per arm, {rec} recruited per arm, {2 * rec} in all")
    print(f"  80 % power alternative: {n_ancova(4.0, 0.5, 1.5, 0.80)} analysable per arm")
    print("sensitivity (analysable per arm, power 90 %):")
    print("   SD    rho    SESOI 1.0   SESOI 1.5   SESOI 2.0")
    for sd in (3.5, 4.0, 4.5):
        for rho in (0.3, 0.5, 0.6):
            print(f"  {sd:.1f}   {rho:.1f}     " + "        ".join(
                f"{n_ancova(sd, rho, d):4d}" for d in (1.0, 1.5, 2.0)))
    rng = random.Random(1954); z = Z(0.975)
    for d in (1.5, 0.4, 0.0):
        hits = sum((lambda e: e[0] - z * e[1] > 0)(trial(n, d, rng)) for _ in range(2000))
        print(f"simulated power, true difference {d:.1f} mm (n = {n}): {hits / 2000:.3f}")
    print("0.4 mm = 0.1 SD, the order of a bias-corrected average focus effect")
# main plan: 113 analysable per arm, 133 recruited per arm, 266 in all
#   80 % power alternative: 84 analysable per arm
# sensitivity (analysable per arm, power 90 %):
#    SD    rho    SESOI 1.0   SESOI 1.5   SESOI 2.0
#   3.5   0.3      235         105          59
#   3.5   0.5      194          86          49
#   3.5   0.6      165          74          42
#   4.0   0.3      306         136          77
#   4.0   0.5      253         113          64
#   4.0   0.6      216          96          54
#   4.5   0.3      388         173          97
#   4.5   0.5      320         142          80
#   4.5   0.6      273         122          69
# simulated power, true difference 1.5 mm (n = 113): 0.896
# simulated power, true difference 0.4 mm (n = 113): 0.130
# simulated power, true difference 0.0 mm (n = 113): 0.024
# 0.4 mm = 0.1 SD, the order of a bias-corrected average focus effect

The sample size rests on the following assumptions (all assumed values, not estimates from any study): an SD of transfer error of 4.0 mm; a correlation of 0.5 between pre-practice error and transfer error (ANCOVA multiplies the variance by 1−0.521 – 0.5^2); a smallest effect of interest of 1.5 mm (about 0.375 SD), two-sided α=0.05\alpha = 0.05 and 90 % power. Each arm needs 113 analysable participants; allowing for about 15 % missing data, 133 are recruited per arm, 266 in all. The 80 % power alternative is 84 analysable per arm. Sensitivity: with an SD of 3.5–4.5 mm and a correlation of 0.3–0.6, 74 to 173 per arm are needed; with a smallest effect of interest of 1.0 mm, 165 to 388. ANCOVA on 2,000 simulated trials gives a power of 0.896, in agreement with the formula; with no true difference, “intervention better” is wrongly concluded in 2.4 %.

It must be said plainly: if the true effect is only 0.4 mm (0.1 SD, the same order as the bias-corrected external-focus effect of Chapter 6), the probability of a significant improvement at this sample size is only 13.0 %. The trial can reliably detect an effect “worth acting on”, not “any effect at all”; if the result is “no difference worth acting on”, that is one of the questions it is meant to answer. Before the main trial a small pilot is run only to check the equipment, the length of the instructions and the difficulty of the task, not to change the sample size.

Decision rule

Calculation

# Decision rule for the Chapter 8 trial, fixed before any data are collected.
# d = adjusted mean transfer RMS error, control arm minus intervention arm (mm; positive
# = the external-focus + manifold-feedback arm is better), with its 95 % CI from the
# ANCOVA. Smallest effect of interest (SESOI) 1.5 mm. Gates -> NOT INTERPRETABLE:
# more than 20 % of randomised participants without a primary outcome; fewer than 70 %
# of either arm passing the instruction check; device calibration failed in > 10 % of
# sessions. Secondary outcomes (incl. ln(Vucm/Vort)) never change the verdict.
import math, random
from statistics import NormalDist
SESOI = 1.5

def verdict(d, lo, hi, missing=0.08, check=(0.90, 0.90), calib=0.02):
    gate = None
    if missing > 0.20:
        gate = f"{100 * missing:.0f} % missing (limit 20 %)"
    elif min(check) < 0.70:
        gate = f"instruction check {100 * min(check):.0f} % (limit 70 %)"
    elif calib > 0.10:
        gate = f"calibration failed in {100 * calib:.0f} % (limit 10 %)"
    if gate:
        return f"d = {d:+.2f}: NOT INTERPRETABLE - {gate}"
    if lo >= SESOI:
        v = "better by at least the SESOI"
    elif lo > 0:
        v = "better; SESOI not shown"
    elif hi < 0:
        v = "WORSE than control"
    elif lo >= -SESOI and hi <= SESOI:
        v = "no difference worth acting on"
    else:
        v = "unclear: interval too wide"
    return f"d = {d:+.2f} mm ({lo:+.2f} to {hi:+.2f}): {v}"

if __name__ == "__main__":
    for case in ((3.2, 1.7, 4.7), (2.0, 0.5, 3.5), (0.3, -1.2, 1.8), (0.2, -1.1, 1.4),
                 (-1.8, -3.2, -0.4), (2.4, 0.9, 3.9, 0.23),
                 (2.4, 0.9, 3.9, 0.08, (0.9, 0.62)),
                 (2.4, 0.9, 3.9, 0.08, (0.9, 0.9), 0.14)):
        print(verdict(*case))
    # Verdict frequencies: 113 per arm, residual SD 4 x sqrt(1 - 0.5^2) mm (the ANCOVA
    # residual), 10,000 simulated trials for each true d.
    rng = random.Random(7); z = NormalDist().inv_cdf(0.975); n = 113
    sd = 4.0 * math.sqrt(0.75)
    short = {"better by at least the SESOI": ">= SESOI",
             "better; SESOI not shown": "better", "WORSE than control": "worse",
             "no difference worth acting on": "none",
             "unclear: interval too wide": "unclear"}
    for true in (1.5, 2.25, 0.4, 0.0):
        tally = {}
        for _ in range(10000):
            a = [rng.gauss(true, sd) for _ in range(n)]
            b = [rng.gauss(0, sd) for _ in range(n)]
            ma, mb = sum(a) / n, sum(b) / n
            ss = sum((x - ma) ** 2 for x in a) + sum((x - mb) ** 2 for x in b)
            d = ma - mb; se = math.sqrt(ss / (2 * n - 2) * 2 / n)
            v = verdict(d, d - z * se, d + z * se).split(": ", 1)[1]
            tally[short[v]] = tally.get(short[v], 0) + 1
        ranked = sorted(tally.items(), key=lambda kv: -kv[1])
        print(f"true d = {true:4.2f}: "
              + ", ".join(f"{k} {c / 100:.1f} %" for k, c in ranked))
# d = +3.20 mm (+1.70 to +4.70): better by at least the SESOI
# d = +2.00 mm (+0.50 to +3.50): better; SESOI not shown
# d = +0.30 mm (-1.20 to +1.80): unclear: interval too wide
# d = +0.20 mm (-1.10 to +1.40): no difference worth acting on
# d = -1.80 mm (-3.20 to -0.40): WORSE than control
# d = +2.40: NOT INTERPRETABLE - 23 % missing (limit 20 %)
# d = +2.40: NOT INTERPRETABLE - instruction check 62 % (limit 70 %)
# d = +2.40: NOT INTERPRETABLE - calibration failed in 14 % (limit 10 %)
# true d = 1.50: better 87.3 %, unclear 7.3 %, none 2.8 %, >= SESOI 2.6 %
# true d = 2.25: better 63.0 %, >= SESOI 36.8 %, unclear 0.1 %, none 0.1 %
# true d = 0.40: none 65.6 %, unclear 21.1 %, better 13.1 %, worse 0.2 %
# true d = 0.00: none 80.2 %, unclear 15.0 %, worse 2.4 %, better 2.4 %

The first eight example results are invented. If more than 20 % of randomised participants have no primary outcome, if fewer than 70 % of either arm pass the instruction-comprehension check, or if device calibration failed in more than 10 % of test sessions, the report says “not interpretable” whatever the result. Only when the lower limit of the 95 % confidence interval for dd is at least 1.5 mm does the report say “the intervention is better by at least 1.5 mm”. Simulation shows that with 113 per arm and a true dd of exactly 1.5 mm, 87.3 % of 10,000 simulated trials conclude “better, but the smallest effect of interest not shown” and 2.6 % reach the strongest conclusion, about 90 % together; with a true dd of 2.25 mm, 36.8 % reach the strongest conclusion; with a true dd of 0.4 mm, 65.6 % conclude “no difference worth acting on”; with no true effect, 80.2 % conclude “no difference worth acting on”, and 2.4 % each wrongly conclude “better” or “worse”. All secondary outcomes (including ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}})) are reported only as estimates with intervals and do not change the primary conclusion.

What this trial cannot show

First, the intervention combines external instructions with redundancy-manifold feedback, and the control combines internal instructions with joint-deviation feedback; a positive result cannot tell whether the instruction or the feedback did the work, which would need a later factorial design. Second, this is a two-joint tracking task in a laboratory, and the result cannot be generalised directly to sport, instrument or singing training, or to rehabilitation; least of all to people with injuries or neurological conditions. Third, the outcome is measured after 24 hours and says nothing about longer retention. Fourth, the secondary outcome ln⁡(VUCM/VORT)\ln(V_{\mathrm{UCM}}/V_{\mathrm{ORT}}) depends on the pre-specified coordinates and units, and other coordinates might give different conclusions (Chapter 2). Fifth, participants know which instruction they received, so expectation effects cannot be fully excluded. Sixth, if the true effect is very small (about 0.1 SD), the trial will most likely conclude “no difference worth acting on”, which does not prove that the two forms of instruction are identical.

Ethics and conflicts of interest

This is a low-risk behavioural study: the task consists of small seated arm movements and involves no medical intervention. Even so, the trial must be reviewed and approved by an independent ethics committee, every participant gives informed consent, and the protocol and analysis plan are registered before the first participant is enrolled (Nosek et al. 2018); the protocol follows the items of SPIRIT 2013 (Chan et al. 2013). The task stops at any pain or discomfort, participants may withdraw at any time and are paid for the part completed, and people with injuries are not enrolled and are advised to consult a professional. Data are stored anonymously and used only to compute the pre-specified measures. After the trial the control arm is offered the intervention arm’s instruction materials. The results will be published whether positive or negative. The trial is run and analysed by researchers independent of the author’s business.

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 teaches singing at vocal.guru and singingasia.com and develops movement and perception training courses, and has a direct commercial interest in the subject of this book. The book promotes no course, card deck or service and is not sports-medicine, physiotherapy or rehabilitation advice; 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 customers; all the code and the study protocol are published openly for anyone to check.

References

Chan, An-Wen, Jennifer M. Tetzlaff, Douglas G. Altman, et al. 2013. “SPIRIT 2013 Statement: Defining Standard Protocol Items for Clinical Trials.” Annals of Internal Medicine 158 (3): 200–207. https://doi.org/10.7326/0003-4819-158-3-201302050-00583.
Chua, Lee-Kuen, Judith Jimenez-Diaz, Rebecca Lewthwaite, Taewon Kim, and Gabriele Wulf. 2021. “Superiority of External Attentional Focus for Motor Performance and Learning: Systematic Reviews and Meta-Analyses.” Psychological Bulletin 147 (6): 618–45. https://doi.org/10.1037/bul0000335.
McKay, Brad, Abbey E. Corson, Jeswende Seedu, et al. 2024. “Reporting Bias, Not External Focus: A Robust Bayesian Meta-Analysis and Systematic Review of the External Focus of Attention Literature.” Psychological Bulletin 150: 1347. https://doi.org/10.1037/bul0000451.
Nosek, Brian A., Charles R. Ebersole, Alexander C. DeHaven, and David T. Mellor. 2018. “The Preregistration Revolution.” Proceedings of the National Academy of Sciences of the United States of America 115 (11): 2600–2606. https://doi.org/10.1073/pnas.1708274114.
Todorov, Emanuel, and Michael I. Jordan. 2002. “Optimal Feedback Control as a Theory of Motor Coordination.” Nature Neuroscience 5 (11): 1226–35. https://doi.org/10.1038/nn963.
Yang, Jeng-Feng, and John P. Scholz. 2005. “Learning a Throwing Task Is Associated with Differential Changes in the Use of Motor Abundance.” Experimental Brain Research 163 (2): 137–58. https://doi.org/10.1007/s00221-004-2149-x.

Fan, Y. (2026). Embodied Cognition, Bernstein's Degrees-of-Freedom Problem and the Control Theory of Motor Synergies: From the Uncontrolled Manifold Hypothesis and Internal Forward Models to Fitts's Law, Read Beside Classical Chinese Thought on Skill and the Way [具身认知、伯恩斯坦自由度问题与运动协同的控制理论] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23828057 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).