‘Adult Status of Children with Contrasting Early Experience’ by Harold M. Skeels

Harold M Skeels (1966) Adult status of children with contrasting early experience: a follow-up study Monographs of the Society for Research in Child Development 31(3):1-65

Harold Manville Skeels (1901-1970) was a psychologist who, following his appointment in 1932 as one of the first psychologists working in Iowa State institutions, became associated with research which undermined the dominant view of intelligence test results. His first paper in 1936 arose out of this work and this paper, one of his last, revisits perhaps his most significant research project in terms of confounding expectations.

Key Ideas

– IQ test results are significantly affected by a person’s current circumstances.

– Quality care is associated with positive one-to-one relationships.

– Women with learning disabilities can care satisfactorily for young children.

– Poor quality care costs five times more than good quality care.


Harold Skeels starts by saying that this study concerns 25 children, 20 of whom were illegitimate, who were admitted before the age of two to the nursery of an orphanage where there was little stimulation and then moved to cottages of 30-35 children where they largely remained inside until the age of six when they started school. From time to time some of the children who appeared less able would be transferred to institutions for mentally disabled children.

In 1932 he had been appointed as psychologist to the home and had transferred two ‘retarded’ children under 18 months old to a home for the mentally disabled. Six months later, visiting the home, he saw two girls whom he did not recognise and was told they were the ones he had transferred. They were performing at a ‘normal level’ and the same was found twelve months later. They had been placed on a ward for 18-50 year old women with mental ages of 5-9; two older inmates had ‘adopted’ each of the girls while the others acted as ‘adoring aunts’. He subsequently arranged for the girls to be moved back to the orphanage and placed for adoption.

The consistent element appeared to have been the one-to-one relationship with a loving and affectionate adult.

By now regular testing of all the children in the orphanage was being undertaken and children identified as unsuitable for adoption often ended up in homes for the mentally disabled. So he proposed that a group of children whose test results suggested they would end up in such institutions be transferred as ‘house guests’ at Glenwood State School, a local institution for the mentally disabled. The children were transferred in groups of three or four, though not all wards that were used had as favourable conditions as the one where the two girls had been.

The experimental group comprised ten girls and three boys; later a contrast group, all of whom had remained in the orphanage until at least the age of four, was selected, based on the comprehensive data that was now available at the orphanage. Though their family backgrounds were similar, several of the experimental group had been premature babies and their IQ test results at the point they were selected for the experiment had been lower than those of the contrast group (Table 1) (Skeels and Dye, 1939).

Table 1: Changes in IQ test results during the experiment

  Starting range Ending range Change first to last test
Experimental 35-89 77-113 +7 ? +58
Contrast 50-103 42-83 -45 ? +2

At Glenwood State School there was competition between the wards over the progress of the children and in nine of the thirteen cases at least one older inmate became attached to a child.

Statistical tests showed there were no associations in the gains/losses of IQ by children to their family history or IQ; the children who made the least progress initially were later placed on another ward where they also made progress.

All thirteen children were eventually returned to the orphanage; one stayed there until adulthood, five were adopted directly, six were adopted later and one was transferred to an institution for the mentally retarded.

Follow up checks were made two and a half years later (Table 2) (Skeels, 1942).

Table 2: Changes in IQ test results after 2½ years

  Follow-up test range Total change
Experimental 63-118 +2 ? +61
Contrast 35-91 -64 ? +18

The adult follow-up began 20 years after the post-experimental follow-up; one of the contrast group had died; six of the experimental group were still living in Iowa, seven had moved outside; nine of the contrast group were still living in Iowa, two had moved outside.

All the experimental group were self-supporting whereas only one of the contrast group was really self-supporting. Table 3 summarises the main results of the follow up study.

Table 3: Adult follow up results

  Experimental Contrast
First study average change in IQ +28.5 average change in IQ -26.2
Follow-up all self-supporting eight in/associated with institutions
Education median 12th grade median less than 3rd grade
Work all in work and/or married to someone in work one in work
Family eleven married; nine have twenty-eight children between them two married; five children between them
Their children 86-125 IQ (median 104) one child with a learning disability; four with normal IQs

The costs of caring for the contrast group had been five times that of caring for the experimental group and the cost of each of the four still in residence in institutions for the mentally disabled will be at least $200/month for the next 20-40 years.

The two key implications of this study are:

– the need for better care;

– you cannot predict IQ on a single test.


Harold Skeels’s work was possibly overlooked because the outbreak of the Second World War had diverted people’s attention in other directions and it was possibly ignored because it was assumed to relate to the unfashionable area of learning disabilities. His research goes to the heart of two 20th century debates: the significance of test results and the nature of good care.

Glenwood State School had evolved out of the Iowa Institution for Feeble Minded Children which had embraced eugenics; accommodating over 1,500 residents, it was to become the Glenwood State Hospital School and take part in the 1950s in the state sponsored sterilisation experiments. In this context, Skeels’s experiment with thirteen children was almost invisible and indeed remained invisible in the UK where Sir Cyril Burt’s fraudulent views (Kamin, 1974) dominated education policy. Though probably unaware of Skeels’s research, Clegg and Megson (1968) are among the few advocates of changing the relationships in children’s environment to improve their life chances.

Skeels shows, as Trotzkey (1930) had argued not many years earlier, that it is not who cares for a child but the quality of the relationship that they have with the child that is the key to good care; all the arguments about the importance of the natural mother or of finding a child a family are irrelevant. What a child needs is a positive relation with someone; it does not even need to be the same someone throughout their childhood as Wiener and Wiener (1990) were to find.

Skeels also demolishes the argument that mothers with a learning disability cannot care for children; while he assumes that, once the child reaches the mental age of the mother, there may be problems (though of course the children in his experiment never reached that age), his research certainly shows that women with learning difficulties are more than capable with appropriate support of providing excellent care for young children.

Finally, Skeels’s research is significant in that, like Trotzkey (1930) and Wiener and Wiener (1990), he focuses on the outcomes for the children and, like Taylor and Alpert (1973) and Fanshel and Shinn (1978), he was not expecting the results he got.


Clegg, A B & Megson, B E (1968) Children in distress Harmondsworth: Penguin

Fanshel, D & Shinn, E B (1978) Children in foster care: a longitudinal investigation Guildford: Columbia University Press See also Children Webmag March 2009

Kamin, L J (1974) The science and politics of IQ Chichester: Wiley

Skeels, H M (1942) A study of the effects of differential stimulation on mentally retarded children: a follow-up report American Journal of Mental Deficiency 46, 340-350

Skeels, H M & Dye, H B (1939) A study of the effects of differential stimulation on mentally retarded children Proceedings and Addresses of the American Association on Mental Deficiency 44(1), 114-136.

Taylor, D & Alpert, S W (1973) Continuity and support: following residential treatment New York: Child Welfare League of America See also Children Webmag March 2009

Trotzkey, E L (1930) Institutional care and placing-out; the place of each in the care of dependent children Chicago: The Marks Nathan Jewish Orphan Home See also Children Webmag November 2008

Wiener, A & Wiener, E (1990) Expanding the options in child placement Lanham MD: University Press of America

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.