Editing Kinsey Reports (section)

=== Statistics ===
In 1948, the same year as the original publication, a committee of the [[American Statistical Association]], including notable statisticians such as [[John Tukey]], condemned the sampling procedure. In a tense meeting with Kinsey, Tukey supposedly declared that even a sample as small as three to five, chosen randomly, would be preferable to hundreds in Kinsey's sample.<ref>Jones 1997, [[Iarchive:alfredckinseypub0000jone/page/658/mode/2up|p. 658]]</ref> In 1954, leading statisticians, including [[William Gemmell Cochran]], [[Frederick Mosteller]], John Tukey, and [[W. O. Jenkins]] issued for the [[American Statistical Association]] a critique of Kinsey's 1948 ''Male'' report, stating:

<blockquote>Critics are justified in their objections that many of the most interesting and provocative statements in the [Kinsey 1948] book are not based on the data presented therein, and it is not made clear to the reader on what evidence the statements are based. Further, the conclusions drawn from data presented in the book are often stated by KPM [Kinsey, Pomeroy, and Martin] in much too bold and confident a manner. Taken cumulatively, these objections amount to saying that much of the writing in the book falls below the level of good scientific writing.<ref>{{Cite book |last1=Cochran |first1=William G |title=Statistical problems of the Kinsey report on sexual behavior in the human male: a report of the American Statistical Association committee to advise the National Research Council, Committee for Research in Problems of Sex |last2=Mosteller |first2=Frederick |last3=Tukey |first3=John W |last4=American Statistical Association |last5=National Research Council (US) |last6=Committee for Research in Problems of Sex |date=1954 |publisher=Greenwood Press |location=Westport, Conn. |language=en |oclc=2908863}}</ref></blockquote>

In response, [[Paul Gebhard]], Kinsey's close colleague, "Sexual Behavior in the Human Female" co-author, and successor as director of the [[Kinsey Institute for Sex Research]],<ref name="nygebhard">{{Cite news |last1=Roberts |first1=Sam |date=2015-07-22 |title=Paul Gebhard, Sex Researcher Who Worked with Kinsey, Dies at 98 |newspaper=The New York Times |url=https://www.nytimes.com/2015/07/23/us/paul-gebhard-sex-researcher-who-worked-with-kinsey-dies-at-98.html |url-status=live |url-access=subscription |archive-url=https://ghostarchive.org/archive/20220103/https://www.nytimes.com/2015/07/23/us/paul-gebhard-sex-researcher-who-worked-with-kinsey-dies-at-98.html |archive-date=2022-01-03}}{{cbignore}}</ref> cleaned the Kinsey data of purported contaminants, removing, for example, all material derived from prison populations in the basic sample.<ref name="gebhardrelia">{{cite web |author=New River Media |title=NEW RIVER MEDIA INTERVIEW WITH: PAUL GEBHARD Colleague of Alfred Kinsey 1946-1956 Former Director of the Kinsey Institute |url=https://www.pbs.org/fmc/interviews/gebhard.htm |access-date=May 27, 2017 |publisher=PBS.org}}</ref> (Gebhard had, while working with Kinsey, raised serious concerns about the use of prison populations especially, but had been shot down by Kinsey at the time.<ref>{{Cite book |last=Gathorne-Hardy |first=Jonathan |title=Sex the Measure of All Things: A Life of Alfred C. Kinsey |publisher=University of Indiana Press |year=2000 |isbn=0-253-33734-8 |location=Bloomington, IN |page=326}}</ref>) In 1979, Gebhard (with Alan B. Johnson) published ''The Kinsey Data: Marginal Tabulations of the 1938–1963 Interviews Conducted by the Institute for Sex Research''. Their conclusion, to Gebhard's surprise, was that none of Kinsey's original estimates were significantly affected by this bias: that is, the prison population and male prostitutes had the same statistical tendencies as the rest of the men Kinsey interviewed. The results were summarized by historian, playwright, and gay-rights activist [[Martin Duberman]]: "Instead of Kinsey's 37% (men who had at least one homosexual experience), Gebhard and Johnson came up with 36.4%; the 10% figure (men who were "more or less exclusively homosexual for at least three years between the ages of 16 and 55"), with prison inmates excluded, came to 9.9% for white, college-educated males and 12.7% for those with less education.<ref name="MDOG">[http://www.kinseyinstitute.org/publications/duberman.html Martin Duberman on Gebhart's "cleaning" of data] {{webarchive|url=https://web.archive.org/web/20090111215816/http://www.kinseyinstitute.org/publications/duberman.html|date=2009-01-11}}</ref> 

Kinsey himself was extremely frustrated by the criticisms of his sampling procedures, because he maintained that there was no way to do a successful study about sex using random probability sampling.<ref name=":0">{{Cite book |last=Gathorne-Hardy |first=Jonathan |title=Sex the Measure of All Things: A Life of Alfred C. Kinsey |publisher=Indiana University Press |year=2000 |isbn=0-253-33734-8 |location=Bloomington, IN |pages=280-281}}</ref> As Kinsey biographer Jonathan Gathorne-Hardy points out, because of the sensitive nature of a sex study, contacting a truly random sample will garner a very high refusal rate--as modern sex studies using random sampling have shown.<ref name=":0" /> If the study is trying to gather information about any sort of population subgroups, as Kinsey was, the small percentages of the population plus the high rates of refusal may make these subgroups effectively disappear, despite their importance to the study.<ref name=":0" />