Theory predicts that parents adjust the sex ratio of their brood to the sexually selected traits of their mate because the reproductive success of sons may be more dependent on inherited paternal attractiveness than that of daughters. Empirical studies vary in terms of whether they support the theory, and this variation has often been regarded as evidence against sex ratio adjustment or has been ascribed to methodological differences. Applying phylogenetic meta-analyses, we aimed to find biological explanations for the variation observed in songbirds. In particular, we tested the role extra-pair paternity, because infidelity occurs in the majority of these species and may reduce the adaptive value of adjusting brood sex ratio to the phenotype of the social mate. However, we found that the variation in effect sizes was unrelated to the proportion of extra-pair paternity. Thus future studies should consider that mate quality dependent sex ratio adjustment may be driven by direct (material) rather than indirect (genetic) benefits. We also tested if the effect sizes are influenced by whether the focal male trait is indeed under sexual selection as it is assumed by the sex allocation theory. We found that for male traits with proven role in sexual selection, effect sizes significantly differed from the null expectation of random production of sons and daughters. For male traits with only presumed sexual role in sexual selection, the deviation from the null expectation was less convincing, and the effect sizes were significantly smaller. This result indicates that studies that neglect the assumptions of the hypotheses concerned, may lead to the underestimation of the mean effect size and, eventually, false conclusions.