The counselling in the Cambridge-Somerville Youth Study lasted six and a half years. The selection that came before it lasted four, and almost everything that makes the study readable eighty years later was built in those four years. This piece follows the machinery, from the study’s own account of it, written by its director in 1940 for the journal Child Development; from a Boston University thesis of 1947 whose author worked in the files with Edwin Powers’s unpublished manuscript beside him; and from the counts in the final report of 1951, as the study’s present directors have reproduced them.

Names

The study was incorporated in the autumn of 1935, as a private social agency of the Ella Lyman Cabot Foundation, and its first act was to get permission from the school committees of Cambridge and Somerville. Then, the 1940 article says, “much work was done in interpreting the objectives” to the schools and the agencies, and “especially great care was taken to prevent the use of such terms ‘delinquent’, ‘behavior problem’, ‘maladjustment’ when referring to any boy or group of boys then or later included in the program.”

Early in the 1935 school year the teachers in the public and parochial schools were asked for the names of boys “who, in their opinion, needed the kind of help the agency was prepared to offer,” and, at the same time, for the names of “average” boys of the same ages and grades, so that the list would carry no stigma. Social agencies, settlements and recreation centres sent names too, and so did the police and the probation officers. The director counted 1,866 boys referred; the 1951 report counted 1,953 screened; the 1947 thesis says “more than nineteen hundred.” About three quarters came from the schools. The age range was five to thirteen, and most were between nine and twelve.

The folder

For each boy the staff then built a folder, and the 1947 thesis lists what went into it: a Stanford Achievement Test; a Kuhlmann-Anderson group intelligence test; a physical examination, with a schedule filled in by the doctor and the nurse recording their impressions of the boy’s personality; a school photograph; a Haggerty-Olson-Wickman behaviour schedule; the Board of Probation’s record, if any, on the boy, his parents and his brothers and sisters; the Social Service Index; summaries of every other agency’s file on the family; the home visitor’s questionnaire, “containing a great deal of detailed information about the boy and parents, obtained in a direct interview with the boy’s mother”; the home visitor’s ratings of the standard of living, of the home “as a place likely to produce delinquency,” and of the boy’s discipline; an interview with his teacher, in which the teacher was asked to predict his future; a descriptive rating scale the teacher filled in; a trait record card, “a check list of ninety-three unfavorable characteristics,” which the teacher ticked and double-ticked, writing a character sketch on the back; ratings by the doctor, the nurse, the home visitor, two psychologists and anyone else on the staff who had met the boy; and a rating of his neighbourhood. The home visitor was a woman, and she explained the programme to the mother “in simple terms” and asked about the boy’s development, his play, his attitudes at home, the family’s money, the parents’ work, schooling, health and personalities, the other children, the street and the house. Every staff member who met a boy or read his papers recorded an impression, independently of the others.

In July 1937 the staff tried something the papers could not do. Forty-seven of the oldest boys were taken on overnight camping trips in groups of between five and ten, under two members of staff, and watched. The trips “helped to establish cordial relationships with the boys’ families, indicating that the C.S.Y.S. was anxious to offer tangible services,” and let the staff revise what they thought they knew. The 1951 report puts the number observed on such trips at eighty. The collecting of information “was not finally completed until the spring of 1939.”

Pre-selection

Only 650 boys were wanted, and nearly three times that number had been named. Boys who had passed their twelfth birthday by the time their data were gathered were dropped, with boys who had moved away or could not be found. By the 1951 count, 538 were dropped because their families did not meet the study’s criteria, and 633 for being too old, gone, unlocatable, insufficiently documented, dead, uncooperative or named twice. When it appeared that too many boys who were not really “pre-delinquent” would get through, the psychologist P. S. deQ. Cabot ran a pre-selection “to eliminate as far as possible the undue proportion of ‘average’ boys,” with the aim of holding the average boys to a third of the total and keeping as many of the youngest pre-delinquents as possible.

The judges

The folders of the boys who remained went to a selection committee of three men who were not on the staff, chosen from thirty-seven candidates, none of whom had gathered the data or made the original ratings. The 1947 thesis gives their posts without their names: the director of case work at the Massachusetts Reformatory, formerly of the state prison; a practising psychiatrist, formerly of the Judge Baker Guidance Center in Boston and of the state hospital at Ann Arbor, the hospital for the criminally insane at Ionia and the Wayne County Training School in Michigan; and the head social worker of the classification division of the Massachusetts Department of Correction, who had been a settlement house worker. The 1940 article calls them “a psychiatrist and two social workers.”

They met twenty-five times between October 1937 and January 1939, “in addition to many preliminary conferences.” Each read every folder on his own, and then the three held “a round table conference on each boy” before deciding; where they could not agree, majority and minority opinions were recorded. They did not interview the boys, and the psychiatrist among them said so: “The lack of any psychiatric interview with the subject did lessen tremendously the accuracy of the work. This, in my opinion, is the most valuable prognostic aid there is.” Each boy was rated on an eleven-point scale from minus five, “the greatest probability that the boy would develop a ‘delinquent career’,” through zero to plus five; a delinquent career was defined as “a customary and habitual mode of adjustment” or “a persistent pattern of anti-social behavior,” and one of the judges held that a court appearance was neither necessary nor sufficient for it. Since the committee did not know which boys would be treated, it was told to predict “on the assumption that the boy would not receive special treatment.” It passed 782: 360 “difficult,” 334 “average,” and 88 in a “zero” group, “about whom there was considerable diagnostic and prognostic doubt.”

Matching

Then a matching committee of three staff members “proceeded to find ‘diagnostic twins’.” Each folder was analysed onto a code data sheet, “so arranged that the data were broken down into approximately 160 separate items grouped under such categories as general and medical information, developmental and personality data, educational and mental status, personality ratings, and final disposition.” The 1951 report describes 142 variables rated on an eleven-point scale. The pairing itself was “a searching clinical review of the code data sheets,” after which the two boys’ figures were set side by side on an analysis sheet under a fixed scheme: chronological age, dominant stock, the birthplace of each parent, religion, locality, school, grade placement, physical health, intelligence, attainment and educational quotients, mental health, social adjustment, the judges’ verdicts, home ratings, standard of living, the father’s occupation, the neighbourhood, the school’s occupational level, and more. “Descriptive reports were made together with a summary of conclusions and reasons for the match.” Six hundred and fifty boys were matched, “of whom 325 constituted the ‘Treatment’ group and a similar number the ‘Control’ group.” The 132 boys left over were “surplus,” in the 1951 report’s word, “beyond the limit determined.”

Each of the two groups was further divided into the difficult, the average and the zero boys, so that, the director wrote, nine comparisons could be made across the groups and more within them: treated difficult boys against untreated difficult boys, and so on.

The coin

The 1940 article says only that after the matching “an arbitrary division of boys into ‘T’ and ‘C’ groups was made.” The 1947 thesis is plainer: “After two boys were paired by the matching process on the basis of I.Q., personality and social background, a coin was tossed to determine which of them would fall into the treatment group and which into the control group.” The 1951 report explains why, in a passage the study’s present directors have quoted: an arbitrary decision “might give rise to a constant error. The proper method of determining this question was, of course, by chance. Accordingly, a coin was flipped and the cases fell into the treatment or comparison groups in accordance with its fall.” Two exceptions were allowed, which Joan McCord later counted. Eight boys were matched after their treatment had begun. And brothers were kept together, in whichever group the first brother had drawn: twenty-one boys in the treatment group and nineteen in the control group owed their place to a brother’s coin rather than their own.

Assignment and the first year

The boys were assigned to counsellors as the pairs were made, “many of whom are trained social workers and others of whom have had considerable experience with adolescents and young boys,” from November 12, 1937 to May 13, 1939, with attention to the boy’s age, school, personality and apparent needs, the counsellor’s qualifications and sex, and where the boy lived. A full-time counsellor carried about thirty-four boys. “June 1, 1939 was arbitrarily set as the date when treatment began after the last case had been assigned.” By the time the director wrote, in 1940, five treated boys had moved beyond the staff’s reach and two had died, and the treatment case load was 318.

For 1939 he printed the year’s work. The counsellors recorded 2,353 interviews with families, 2,717 with boys, 1,119 with schools, 521 with social agencies, 236 with settlements, 485 with hospitals, 117 with the delinquency squad and the police, 59 with probation officers, 23 at court, 577 among the staff and 276 miscellaneous: 8,483 in all. The services of the year included psychological testing, remedial work, psychiatric referrals, school conferences, recreation, trips, special training, hobbies and camps, help with family problems, cash, clothing, advice, employment, personal adjustment, health examinations, treatment plans, dental examinations and dental treatment, and foster-home placements. Eighty-eight cooperating agencies in greater Boston provided what the study could not. The boys were in twenty-one public and ten parochial schools in Cambridge, 356 and 65 of them, and in the public and parochial schools of Somerville, with thirty-six in ten other school systems in Massachusetts to which families had moved. The director held a monthly case conference with each counsellor; a casework committee of three social workers from other Boston agencies met monthly; and a case consultant was on call.

The staff in 1940 was a director, nine counsellors, an associate counsellor, a casework consultant, two psychologists, four volunteer tutors, a supervisor of controls, a secretary and five secretarial staff including a research assistant. The supervisor of controls collected, on the untreated boys, their school progress, any apprehended delinquency, their agency contacts, special tests and medical examinations, through teachers, probation officers, clergy, agencies, camp directors and “members of the treatment staff whose ‘T’ boys may be friends of ‘C’ boys.”

The general advisory committee listed in 1940 was William Boyd of Glasgow; Stanley Cobb of the Harvard Medical School; Donald MacPherson, psychoanalyst; Henry Murray of the Harvard Psychological Clinic; Jean Piaget of Geneva; the paediatrician E. C. Romberg; A. Warren Stearns, dean of Tufts Medical School, whose words Cabot had argued with in 1931; D. A. Thom of the West End Habit Clinic; Gordon Allport; Edith Baylor of the Children’s Aid Association; Phillip Rulon of the Harvard school of education; Hadley Cantril of Princeton; and the anthropologist Earnest Hooton. The treatment advisory committee was Baylor, Monsignor Augustine Hickey of St Paul’s, Cambridge, Augusta Bronner and William Healy of the Judge Baker Guidance Center, Cheney Jones of the New England Home for Little Wanderers, Theodore Lothrop of the Massachusetts Society for the Prevention of Cruelty to Children, and Alfred Whitman of the Children’s Aid Association.

The director’s last paragraph on research says that cross-sectional analyses of the treated and untreated boys would be made “at periods of no less than two or no more than three years within the originally planned ten-year period,” so that “the final labors of evaluation should be considerably reduced.” The ten years became six and a half; the final labours fell to Helen Witmer in 1948, to the McCords in 1956, and to Joan McCord from 1975.

Sources

• P. S. deQ. Cabot, “A Long-Term Study of Children: The Cambridge-Somerville Youth Study,” Child Development 11, no. 2, June 1940, pp. 143 to 151. The six stages, the 1,866 referrals, the camping trips, the selection committee’s meetings and its 782 boys, the 160 items and the matching variables, “diagnostic twins,” the assignment dates, June 1, 1939, the 1939 tables, the schools, the staff and the committees. archive.org
• Donald Douglas Dowling, An Analysis of the Outcome of Predictions of Delinquency Based on One Hundred Cases of the Cambridge-Somerville Youth Study for the Period 1938 to 1947, thesis, Boston University, 1947, chapters I and II. The contents of the folder, the pre-selection by P. S. deQ. Cabot, the three judges and their posts, the thirty-seven candidates, the eleven-point scale and its definitions, the psychiatrist’s remark, and the coin. archive.org
• Brandon C. Welsh, Steven N. Zane, Gregory M. Zimmerman and Alexis Yohros, “Association of a Crime Prevention Program for Boys With Mortality 72 Years After the Intervention,” JAMA Network Open 2, no. 3, 2019, e190782, for the 1951 report’s counts of 1,953 screened, 538 and 633 excluded, 132 surplus, 142 variables and eighty boys on camping trips, and the two exceptions to the coin. doi:10.1001/jamanetworkopen.2019.0782
• Brandon C. Welsh, Scott H. Podolsky and Steven N. Zane, “Richard Cabot, pair-matched random allocation, and the attempt to compare like with like in the social sciences and medicine. Part 1,” Journal of the Royal Society of Medicine 114, 2021, pp. 212 to 217, for the passage of Powers and Witmer on the coin. doi:10.1177/01410768211003771
• Joan McCord, “Some Unanticipated Consequences of Summer Camps,” 1997, footnote 1, for the eight late matches and the forty brothers. ERIC ED412481