“…When conducting family‐level analyses, I tried to overcome some long‐standing, and largely overlooked, conceptual and methodological problems in sex‐ratio research. Several earlier studies on raptors have investigated correlations between brood sex ratios and food resources, but their results were often difficult to interpret, because: (1) they measured environmental prey abundance, not realized resource use; and (2) they integrated prey abundance over large areas, typically the entire study plot (Bednarz & Hayden, 1991; Wiebe & Bortolotti, 1992; Dzus, Bortolotti & Gerrard, 1996; Leroux & Bretagnolle, 1996; Appleby et al ., 1997; Korpimäki et al ., 2000; Byholm et al ., 2002a; Hipkiss & Hörnfeldt, 2004; Laaksonen, Lyytinen & Korpimäki, 2005; Millon & Bretagnolle, 2005; Desfor, Boomsma & Sunde, 2007). While some of these studies were concerned with population‐level patterns, others aimed to examine individual‐level responses, violating critical theoretical assumptions.…”