What does Attachment and Couple Dynamics conclude, in numbers, about “Is Early Experience Destiny? The Real Effect Sizes of Longitudinal Studies”?
Answer: The average stability correlation of attachment security from infancy to early adulthood is about 0.39, explaining about 15 % of the variance, and over intervals longer than 15 years it is no longer significant. Early care predicts adult attachment, but with small effects. Attachment styles change; a partner's way of responding may be one source of change, but causal evidence is still scarce; retrospective "earned security" may be affected by depression-related recall bias. Early experience matters, but it is not destiny. The next chapter returns to conflict itself: what makes it stop.
Scope of the numbers: Marked boxes separate published evidence, runnable Python calculations (14 programs, all checked), classical sources and the author's hypotheses.
Data table
The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 0.39, 15
1. r = 0.39 means the earlier measure accounts for r^2 = 15% of the variance of the later one 2. same side of a median split both times: 63% (50 % by chance alone; kappa = 0.26) 3. correlation between age 1 and later ages under two models years later | 1 | 2 | 5 | 10 | 15 | 20 revisionist | 0.80 | 0.64 | 0.33 | 0.11 | 0.04 | 0.01 prototype | 0.85 | 0.73 | 0.50 | 0.33 | 0.28 | 0.26 the revisionist model decays toward 0; the prototype model levels off at the stable share (0.25)
Show the full program
# How much does early attachment fix adult attachment? Pinquart, Feussner & Ahnert (2013) pooled
# 127 studies: overall stability r = .39, higher over short intervals, and no significant stability
# over intervals longer than 15 years. Three ways to read such numbers.
import math
r = 0.39
print(f"1. r = {r:.2f} means the earlier measure accounts for r^2 = {r * r:.0%} of the variance of the later one")
# 2. If 'secure' vs 'insecure' is a median split of an underlying continuous variable, the chance
# that a person falls on the same side twice is 1/2 + arcsin(r)/pi (bivariate normal, Sheppard).
same = 0.5 + math.asin(r) / math.pi
print(f"2. same side of a median split both times: {same:.0%} (50 % by chance alone; kappa = {2 * same - 1:.2f})")
# 3. Two simple models of change (compared by Fraley 2002). Yearly correlation of the changing part
# 0.80 in both; the 'prototype' model adds a stable component carrying 25 % of the variance.
print("3. correlation between age 1 and later ages under two models")
print(" years later | 1 | 2 | 5 | 10 | 15 | 20")
rev = [0.8 ** k for k in (1, 2, 5, 10, 15, 20)]
pro = [0.25 + 0.75 * 0.8 ** k for k in (1, 2, 5, 10, 15, 20)]
print(" revisionist | " + " | ".join(f"{v:.2f}" for v in rev))
print(" prototype | " + " | ".join(f"{v:.2f}" for v in pro))
print(" the revisionist model decays toward 0; the prototype model levels off at the stable share (0.25)")Single source and citation
Fan, Y. (2026). Computational Attachment Theory and Couple Interaction Dynamics (Version 1.0.0), Chapter 5: Is Early Experience Destiny? The Real Effect Sizes of Longitudinal Studies. 5D Library. https://doi.org/10.5281/zenodo.23754715
DOI: 10.5281/zenodo.23754715 · Read the monograph · 5DL-B-021 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026021,
author = {Fan, Yafu},
title = {Computational Attachment Theory and Couple Interaction Dynamics},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
doi = {10.5281/zenodo.23754715},
url = {https://doi.org/10.5281/zenodo.23754715}
}