“…As a result, we eliminated a number of studies from review because they failed to meet one or more criteria. These included studies that covered only congressional (e.g., Wells & King, 1994;Coffey, 1975;Kelley, 1958) or gubernatorial (e.g., Ostroff & Sardell, 1989;Bishop & Brown, 1967) races; that examined structural elements of the campaign coverage without making a party-by-party breakdown (e.g., Noyes, Lichter, & Amundson, 1993;Russonello & Wolf, 1979;Patterson, 1989); or that considered only one candidate (e.g., Johnson, 1997;Merron & Gaddy, 1986). One omitted study (Just, Crigler, Alger, Cook, Kern, & West, 1996) yielded only aggregate data from February through November (creating a proincumbent skew).…”