Actual legislators show extreme partisan differences in authoritarianism levels

Evidence

In the Manitoba legislature, the Tory sample mean was twice as high as that of the NDP (no Liberals sat in the Legislative Assembly at the time), with no overlap in the distributions. The point-biserial correlation was but a trifling .91. Some overlap did appear in the Ontario assembly, but the NDP-Liberal- Conservative means lined up as expected and eta-squared equaled 43%, analogous to a correlation of .66. British Columbia (then with no Liberals) resembled Manitoba, with the Social Credit caucus scoring twice as high as the NDPers, with no overlap, and the correlation a noteworthy .85. The New Brunswick House produced the familiar NDP-Liberal-Conservative order, with appreciable overlap between Liberal and Tory camps. Eta-squared proved 29%, corresponding to an r of .54. — Bob Altemeyer, The Authoritarian Specter, 510

Categorization

• XREF: Lends empirical weight to claims about partisan polarization in personality — extends vault material on who wields political authority.
• CANDY: Canadian Tory and NDP caucuses showed zero overlap in authoritarianism scores — a memorable political fact.