The Southwestern Historical Quarterly, Volume 107, July 2003 - April, 2004

136 Southwestern Historical Quarterly July
Misspecification of the precise theoretical model to be tested, lack of any
"weighting" formula for units of analysis, improper operationalization of vari-
ables, and failure to calculate standardized regression coefficients, are among
the numerous problems that vitiate the author's analysis. Bushels of wheat, for
example, are negatively associated with secession, but this relationship is
ignored in the rush to demonstrate that wheat growing had "a significant rela-
tionship to the support for secession" (p. 238) and thus was one of many eco-
nomic interests to be protected politically. If the relative impact of one variable
(i.e., a raw number of bushels of rye) over a totally different variable (i.e., dol-
lar values of slain animals) is being measured to explain the variation in an
independent variable, then one must examine standardized regression coeffi-
cients or beta coefficients. Alas, none are to be found. Allowing the smallest
populated counties to be weighted in the equations on an equal basis with the
largest and failing to provide an indexing of the relative importance of farm
production in a county, predictable lead to absurd conclusions. For example,
the percentage of ballots cast for secession (a variable that ignores turnout rates
of voters) is regressed upon raw numbers of ginned cotton bales and slaves-
two variables that are highly associated with one another (a fatal problem called
multicollinearity). But the author reports this flawed result as demonstrating
that "both cotton and slavery cease to have a significant relationship with seces-
sion" and deduces that "the commercial interests associated with slavery are not
as vital as many historians have previously believed" (p. 66). Such a conclusion
is worse than a woefully defective employment of multiple regression analysis, it
is also bad history.
Texas A&M University DALE BAUM
Queen of the Confederacy: The Innocent Deceits of Lucy Holcombe Pzckens. By Elizabeth
Wittenmyer Lewis. (Denton: University of North Texas Press, 2002. Pp. 288,
25 photos, 11 maps, genealogy, notes, bibliography, index. ISBN 1-57441-
146-2. $24.95, cloth.)
Elizabeth Wittenmyer Lewis has written a lively and engaging account of Lucy
Holcombe Pickens, a woman fiercely committed to the Confederate cause, and
the only woman to have her likeness appear on Confederate currency. Born in
Tennessee in 1832, Lucy was the classic Southern belle: beautiful, coy, clever,
and accomplished. Readers will be fascinated by this extraordinarily self-willed
woman who demanded the right to live life on her own terms.
Despite her considerable feminine charm, Lucy appeared to care little for the
conventional domestic rewards of "true womanhood." In marrying Francis
Wilkinson Pickens-who was twice her age and "reputed to be the wealthiest
man in South Carolina" (p. 65)-she exhibited less interest in romantic love
and motherhood than in assuring her own wealth and status. Although Lewis
emphasizes Lucy's loyalty to Francis, her evidence suggests that Francis's death
lifted from Lucy's shoulders the burden of a demanding and jealous husband,
one for whom she had refused to bear a second child.

Texas State Historical Association. The Southwestern Historical Quarterly, Volume 107, July 2003 - April, 2004. Austin, Texas. The Portal to Texas History. http://texashistory.unt.edu/ark:/67531/metapth101224/. Accessed September 22, 2014.