Marianne De Wolff

Who they are

The question De Wolff and Marinus van IJzendoorn set was narrow and important. Ainsworth had proposed that a caregiver's sensitivity, meaning noticing a baby's signal, reading it correctly and responding promptly, is what produces a secure attachment. Thirty years of individual studies had reported effects of wildly different sizes, some of them large enough to be implausible.

They gathered sixty-six studies covering more than four thousand families and separated out the ones that could answer the question cleanly: non-clinical samples, sensitivity measured by observation rather than report, and measured before or alongside rather than after the classification. Twenty-one studies qualified.

Across those, sensitivity accounted for about six parts in a hundred of the differences in security. Their own description is moderately strong, and they add the sentence this site keeps returning to: sensitivity is an important but not exclusive condition of attachment security.

Two further results deserve more attention than they get. Other parenting dimensions, such as cooperation and stimulation, were weaker but not trivially so, which means sensitivity is not uniquely privileged among the things a caregiver does. And the authors anticipate the objection that a moderate correlation is a small basis for a causal claim, arguing that a modest association can still index a powerful process. That is a reasonable argument and it is an argument, not a finding. The intervention trials are what turned it into one.

Key papers

Key papers

De Wolff, M. S., & van IJzendoorn, M. H. (1997). Child Development, 68(4), 571-591. First author, and the paper is routinely miscited to her co-author. Sixty-six studies covering more than four thousand families were gathered, and the ones that could answer the question cleanly were separated out: non-clinical samples, sensitivity measured by observation rather than by report, and measured before or alongside rather than after the attachment classification. Twenty-one studies qualified. Across those, caregiver sensitivity accounted for about six parts in a hundred of the differences in security. The authors describe that as moderately strong and add the qualification this site keeps returning to, that sensitivity is an important but not exclusive condition of attachment security. Other parenting dimensions, cooperation and stimulation among them, were weaker but not trivial, so sensitivity is not uniquely privileged among the things a caregiver does.

Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Psychological Bulletin, 129(2), 195-215. The randomised intervention meta-analysis, and the reason the 1997 association can now be read causally. De Wolff and van IJzendoorn argued in 1997 that a modest association can still index a powerful process. That was an argument rather than a finding, and this is the work that turned it into one.

Key ideas

Six parts in a hundred, called moderately strong. Honest labelling of a real but modest effect.

Important but not exclusive. The phrase that keeps the parenting claim proportionate.

Sensitivity is not the only thing that matters, which the popular version tends to forget.

Concepts on this site

Caregiver Sensitivity

---