29 August 2001

Diana Baumrind, Ph.D.
Research Psychologist
Institute of Human Development
University of California Berkeley
Fax: 1-510-642-7969

Dear Dr. Baumrind:

I am writing to express my concern about the paper you presented last week at the annual meeting of the American Psychological Association. I am a psychologist myself, but was unable to attend APA this year. I e-mailed your office a week ago requesting a copy of the study you presented, by e-mail or snail mail, and have received no reply. But I have spoken to a psychologist in your audience who received the handout and took notes, and of course no one in America could escape reading about your study in various newspapers or hearing about it on news shows. So, I have not had the opportunity to yet read the full study and examine your statistical methods in detail, but I believe I have a fairly good read on the research design, and would be happy to have you correct me if I have any misconceptions. And I still await receiving a copy of the study.

I write in hopes that you will immediately issue press releases recanting the conclusions you have been quoted as reaching. You are quoted as saying that mild to moderate spanking is not harmful. Yet your research was based on case studies of only 3 children who had non-spanking parents, and 79 whose parents spanked in the degree you considered mild or moderate. To have only 3 subjects in one of your principal variable groups does not constitute research, does it? I was taught that such an incredibly small number would just constitute an anecdotal report, an observation, not science. If in even my first year of graduate school I had written up a research design with such a small number of subjects, I would surely have received an "F" on the paper, and the quality of my undergrad education would have been in question. As you surely understand, when the number in one group is so tiny as this, huge differences in behavior or cognition would have been necessary to report any significant differences. [And even the spanked kids were grouped into statistical cells of only 6 to 8 children, correct?]

In other words, using only 3 non-spanked kids had the effect of building into the conclusions an impossibility of finding any differences with spanked kids -- right? Don't you agree? I think you would find it impossible yourself to have any professional journal in our field accept such inadequate research for publication. And to state conclusions that have such far reaching effects on the health and safety of children based on a "study" that would not pass peer review is not responsible. Again, if I am incorrect because I have not received the requested paper, please let me know.

You can gain back the professional respect you have lost among your colleagues if you will immediately notify the news media in strong, clear, unambiguous language that the conclusions drawn were incorrect for such limited data, that you erred and wish to have the mistake corrected.

I would be happy to assist you in making this correction and apology widely known and respectfully received.

Looking forward to your response,

Robert E. Fathman, Ph.D.