Reports an error in "Meta-analyses and p-curves support robust cycle shifts in women's mate preferences: Reply to Wood and Carden (2014) and Harris, Pashler, and Mickes (2014)" by Kelly Gildersleeve, Martie G. Haselton and Melissa R. Fales (Psychological Bulletin, 2014[Sep], Vol 140[5], 1272-1280). In the article, all p-curve analyses examining the Context Moderation Hypothesis Prediction mistakenly included the p-value from Little, Jones, Burt, & Perrett (2007) Study 2 for the simple effect of fertility on attraction to facial symmetry in a short-term relationship context (p < .001). The analyses should have instead included the p-value for the fertility X relationship context interaction (p = .011). In addition, the p-curve analyses examining exact two-tailed p-values for the Cycle Shift Prediction should have included an additional p-value from Provost et al. (2008) Study 1 for the main effect of fertility on attraction to gait masculinity. The reported p-value for this effect was .05, making it ineligible for inclusion in p-curves of reported p-values. However, the exact recalculated two-tailed p-value was .049, making it eligible for inclusion in p-curves of exact p-values. The corrected p-curve of exact two-tailed p-values evaluating the Cycle Shift Prediction and Context Moderation Prediction (displayed in Figure 2) now includes a total of 15 p-values (N = 1442) is no longer significantly right skewed χ²(30) = 41.25, p = .08. The corrected p-curve of exact two-tailed p-values evaluating the Cycle Shift Prediction, Context Moderation Prediction, and Partner Qualities Moderation Prediction (displayed in Figure 3) now includes a total of 21 p-values (N = 1707) and continues to be significantly right skewed Χ²(42) = 69.83, p = .004. As part of this correction, the online supplemental materials have been updated. (The following abstract of the original article appeared in record 2014-35938-003.) Two meta-analyses evaluated shifts across the ovulatory cycle in women's mate preferences but reported very different findings. In this journal, we reported robust evidence for the pattern of cycle shifts predicted by the ovulatory shift hypothesis (Gildersleeve, Haselton, & Fales, 2014). However, Wood, Kressel, Joshi, and Louie (2014) claimed an absence of compelling support for this hypothesis and asserted that the few significant cycle shifts they observed were false positives resulting from publication bias, p-hacking, or other research artifacts. How could 2 meta-analyses of the same literature reach such different conclusions? We reanalyzed the data compiled by Wood et al. These analyses revealed problems in Wood et al.'s meta-analysis-some of which are reproduced in Wood and Carden's (2014) comment in the current issue of this journal-that led them to overlook clear evidence for the ovulatory shift hypothesis in their own set of effects. In addition, we present right-skewed p-curves that directly contradict speculations by Wood et al.; Wood and Carden; and Harris, Pashler, and Mickes (2014) that supportive findings in the cycle shift literature are false positives. Therefore, evidence from both of the meta-analyses and the p-curves strongly supports genuine, robust effects consistent with the ovulatory shift hypothesis and contradicts claims that these effects merely reflect publication bias, p-hacking, or other research artifacts. Unfounded speculations about p-hacking distort the research record and risk unfairly damaging researchers' reputations; they should therefore be made only on the basis of firm evidence. (PsycINFO Database Record

中文翻译：

---

“元分析和P曲线支持女性择偶偏好的强劲周期变化：对Wood和Carden（2014）以及Harris，Pashler和Mickes（2014）的答复”：对Gildersleeve，Haselton和Fales的修正（2014）。

在Kelly Gildersleeve，Martie G. Haselton和Melissa的“元分析和p曲线支持女性伴侣偏好的强劲周期变化：回复Wood和Carden（2014）以及Harris，Pashler和Mickes（2014）”中报告了一个错误。 R. Fales（心理公报，2014年9月，第140卷[5]，1272-1280）。在本文中，所有检查情景调节假说预测的p曲线分析都错误地包含了Little，Jones，Burt和Perrett（2007）研究2中的p值，该值是生育力在短时间内吸引面部对称的简单影响。术语关系上下文（p <.001）。相反，分析应包括生育力X关系上下文交互作用的p值（p = .011）。此外，检验准确的两尾p值以进行循环移位预测的p曲线分析应包括Provost等人的其他p值。（2008）研究1关于生育力对步态男性气质的吸引力的主要影响。所报告的此效果的p值为.05，使其不符合所报告p值的p曲线的要求。但是，精确重新计算的两尾p值为.049，使其有资格包含在精确p值的p曲线中。现在，精确的两尾p值的校正p曲线（评估周期偏移预测和上下文缓和预测）（如图2所示）总共包含15个p值（N = 1442），不再显着右偏χ²（ 30）= 41.25，p = .08。精确的两尾p值的校正后的p曲线，用于评估循环移位预测，上下文缓和预测，和合作伙伴质量审核预测（如图3所示）现在总共包括21个p值（N = 1707），并且继续明显偏右Χ²（42）= 69.83，p = .004。作为更正的一部分，在线补充材料已更新。（原始文章的以下摘要出现在记录2014-35938-003中。）两项荟萃分析评估了女性伴侣的排卵周期中排卵周期的变化，但发现的结果却大不相同。在该期刊中，我们报道了由排卵期假说预测的周期转变模式的有力证据（Gildersleeve，Haselton和Fales，2014年）。但是，伍德，克雷索尔，乔希 和Louie（2014）声称缺乏对这一假设的令人信服的支持，并断言他们观察到的少数重要周期变化是由出版偏见，p-hacking或其他研究产物导致的假阳性。同一文献的两次荟萃分析如何得出这样不同的结论？我们重新分析了伍德等人汇编的数据。这些分析揭示了伍德（Wood）等人的荟萃分析中的问题-一些问题已在本期杂志的伍德和卡登（Wood）和卡登（Carden）（2014）的评论中重现-导致他们在其排卵期假说中忽略了明确的证据自己的一套效果。另外，我们提出了右偏的p曲线，它们与Wood等人的推测直接矛盾。伍德和卡登；哈里斯和帕什勒 和Mickes（2014）认为周期变化文献中的支持性发现是假阳性。因此，来自荟萃分析和p曲线的证据强烈支持与排卵假说一致的真实，有力的影响，并且与声称这些影响仅反映出版偏见，p hacking或其他研究结果的说法相矛盾。关于p-hacking的毫无根据的猜测扭曲了研究记录，并冒着不公平地损害研究人员声誉的风险；因此，仅应在有确凿证据的基础上进行核对。（PsycINFO数据库记录 或其他研究文物。关于p-hacking的毫无根据的猜测扭曲了研究记录，并冒着不公平地损害研究人员声誉的风险；因此，仅应在有确凿证据的基础上进行核对。（PsycINFO数据库记录 或其他研究文物。关于p-hacking的毫无根据的猜测扭曲了研究记录，并冒着不公平地损害研究人员声誉的风险；因此，仅应在有确凿证据的基础上进行核对。（PsycINFO数据库记录