Libel in Fact: When are Data Good Enough?
Were the letters from psychiatrists published in Fact magazine real?
Posted November 1, 2009
Researchers and scholars often face this question. I face it as I examine responses from the 1964 Fact survey of psychiatrists.
Last week I described some "wild analysis" on the part of several psychiatrists who responded to the poll.
The psychiatrists were evaluating then-Senator Barry Goldwater's personality. The term, "wild analysis" refers to an evaluation of someone's personality based on unsubstantiated theory and/or implausible reasoning.
Mr. Warren Boroson, the former managing editor of Fact at the time commented on last week's post:
"The person who edited the letters from psychiatrists--not me--did not have an academic background, and condensed some comments and merged others. Which I relate in the interest of historical accuracy."
Fact magazine's publisher, Mr. Ginzburg, acknowledged in court during the libel trial against him and his magazine, that such editing took place.
In his book, Errors, Lies, and Libel, Professor Peter Kane describes how during the last two days of Senator Goldwater's case, the jury heard evidence against Mr. Ginzburg from the pretrial collection of materials and interviews with key individuals (the discovery process). During those two days, Kane opines:
"...the most damaging evidence was the contrast between materials in the questionnaires that had been returned to Fact and the quotations that appeared in the magazine. There were numerous phrases taken out of context, separate responses combined to appear as one, anonymous statements appearing to have identifiable sources, and even statements changed by Ginzburg to enhance or produce negative effects..."
Such editing casts uncertainty over the valid reporting of individual psychiatric opinion in the survey. Increasing the negativity of a reported opinion represents neither good journalistic nor good survey methods (although it can be a thin line between editing to increase ease of reading, and editing to increase impact: mostly negative impact, in this case). From a legal standpoint, such editing established the possibility of malice.
That granted, the words were still those (mostly) of the psychiatrists. There were no accusations that substantial material was added to the original letters. Moreover, although some material was edited, many letters were reprinted as written, and others retained their original character: The five psychiatrists questioned in pre-trial depositions all said that their published opinions were accurate representations of their evaluations of Goldwater.
So the data are imperfect, and analyses should take that into account: An analysis that drew conclusions about the negativity of psychiatrists regarding Goldwater overall, or that drew conclusions about the quality of psychiatrists' writing styles, would be highly questionable.
Other analyses, however, that concern judgments of personality expressed by the psychiatrists of the time may be on firmer footing. The examination of connected sentences within a paragraph is probably worth carrying out as well. If multiple examples of a phenomena (such as "wild analysis") were present, they probably reflect something useful of the original material. Nonetheless, the unasked-for editing of the psychiatrists' responses is a reminder of the potential fallibility of all data.
Professor David Funder of the University of California, Riverside, has a set of informal laws about personality and how it is studied. Funder's third law concerns the collection of data. It states: "Something beats nothing, two times out of three." As he elaborates elsewhere: "The only alternative to gathering information that might be misleading is to gather no information. That is not progress."
The Fact survey and subsequent libel trial involved eminent political and scientific figures, over one thousand psychiatrists, important legal outcomes, new ethical standards, and a colorful historical story, all centered around the various ways personality is judged.
The psychiatrists' letters represent imperfect data as published - that is certain.
Because I believe the case to be important, I will proceed to examine the letters further albeit with some reservations. I'll take Funder's two-to-three odds that the "something" (the survey result) is more informative than nothing, and that thinking about personality judgments in the context of the trial represents progress that is admittedly uncertain and unknown right now, but that may become clearer later.
The quote from Mr. Boroson is from his comment of October 29th, 2009, on last week's post.
The quote "...the most damaging evidence..." and the testimony of the five psychiatrists were both from pp. 47-48 from Kane, P. E. (1992). Errors, lies, and libel. Carbondale and Edwardsville: Southern IIllinois University Press. I capitalized "Fact" in the quote as reproduced here, although it was lowercase in the original.
Funder's third law, and the quote "The only alternative..." are both from p. 21 of Funder, D. C. (2004). The personality puzzle (3rd ed.). New York: Norton.
Copyright © 2009 John D. Mayer