A Second Opinion

        Judith Wallerstein
        The Unexpected Legacy of Divorce

   In 1971 Judith Wallerstein contacted 60 families who recently filed for divorce in Marin County, California.  Marin County is a wealthy, insular area adjacent to San Francisco.  Twenty-five years later, Wallerstein was still in contact with 45 of these families.

   From this group, Judith attempted to draw inferences about the effects of divorce on children.

Invalid Use of Statistics – Skewed Population

   There are two initial problems.  First, the study group is skewed.  Wealthy Marin County is representative of the population of the United States in the same way a county in rural Mississippi is representative of the population of the United States.  In addition, the sample size—45—is too small to make statistically valid generalizations.

   The next problem is Judith compares children from intact families with children of divorce.  To make the comparison valid she needs two groups: a group staying together and a group with similar characteristics divorcing.  But she has only one group—a divorcing group.  There is no control group.

   Let's take a simplified example.  Imagine Judith had the two groups she needed.  In the U.S. about 20% of families involve an alcoholic parent, and at least 15% involve sexual abuse of one or more children in the family.

   At the beginning of her study, 12 families in each group would involve alcoholism, and 9 families involve sexual abuse of children.  If she compared what happened to the children who escaped sexual abuse or an alcoholic home, to the children raised in a sexually abusive or alcoholic home, she could draw some conclusions.

Flawed Methodology

   She could suggest an answer to the question, What would the children of divorce be like if the parents had stayed together?  Unfortunately, since Judith lacks this data, she cannot make valid inferences. 

   Judith Wallerstein lumps families staying together in one group, which she then compares to families divorcing.  This is an apples and oranges comparison.  The first group contains well-married couples as well as problem marriages, while the second group contains only problem marriages.

   There is an inexcusable omission in The Unexpected Legacy of Divorce.

   In the appendix to an earlier book, Surviving the Breakup, Wallerstein admits 50%  of the divorced men and almost half of the women in her study were clinically depressed or suicidal.  They had problems controlling rage, sexual impulses, and dealing with others. 

   An additional 15 to 20% of the men and women were severely troubled individuals.  That includes histories of mental illness, paranoid thinking, bizarre behavior, and manic depressive illness. 

   When Judith Wallerstein talks about the effects of divorce on children, she fails to mention she is not using a random sample of divorced people, but a group of divorced people with severe mental health issues.

   The Wallerstein "study" is an example of statistically flawed research and improper methodology.

   It does not meet accepted standards of scientific research.  

   Judith's thesis is "… that compared to children from intact families, children of divorce follow a different trajectory for growing up.  It takes them longer.  Their adolescence is protracted and their entry into adulthood is delayed." (p. 37)  But this is clearly the case with the children from intact problem marriages as well.

   To try to salvage the study since she lacks a control group, Judith introduces a "comparison group" to match against the divorcing group.  However, this comparison group has a divorce rate only 25% that of nationally selected random sample, which means that on the face of it the comparison is invalid.  (p. 320)

   The decision to divorce or stay married is a serious decision.  Wallerstein's book does not contain the statistically reliable and valid information needed to make a decision one way or the other. 

   Put another way, would the Federal Drug Administration approve use of a medical procedure based on a study like Wallerstein's?  No.  Her study emerges not as science, but as a curiosity.

(Source: The Unexpected Legacy of Divorce by Judith Wallerstein, Julia Lewis, and Sandra Blakeslee. Hyperion, 2000.)