Reward Window Performance (Section 3.1, p. 9). Infants were overall likely to fixate the reward when it appeared on the screen above chance levels (chance = .5). Results of the Linear Mixed Effects model yielded significant main effects of Phase and Block, as well as significant Group by Phase and Group by Condition interactions. Overall, infants fixated the reward more in the pre-switch than the post-switch phase in the monolingual, ß = -.008, SE = .001, CI[-.001, -.006], t = 6.019, p < .001, and bilingual groups, ß = -.004, SE = .002, CI[-.007, -.006], t = 2.354, p = .018, but bilinguals were more likely to fixate the reward in the auditory than the visual condition, ß = .004, SE = .001, CI[.001, .007], t = 2.578, p = .010.

Visual condition (Section 3.2.2, p. 12). Similar to the auditory condition, this model yielded main effects of Block and Phase. Infants’ performance also increased as the task progressed, and infants fixated the correct location in anticipation of the reward to a greater extent in the pre-switch than the post-switch phase, ß = -.085, SE = .035, CI[-.15, -.02], t = 2.416, p = .016. The Group by Phase interaction was also significant. Planned pairwise comparisons revealed that monolingual and bilingual infants performed similarly in the pre-switch phase, ß = .075, SE = .050, CI[-.02, .17], t = 1.493, p = .136, but bilinguals outperformed monolinguals in the post-switch phase, ß = -.115, SE = .049, CI[-.21, -.02], t = 2.317, p = .021. In this task, bilinguals’ performance did not differ between the pre- and post-switch phases, ß = .010, SE = .053, CI[-.10, .11], t = 0.181, p = .856, but monolinguals’ performance was significantly lower post-switch compared to pre-switch, ß = -.179, SE = .046, CI[-.27, -.09], t = 3.909, p < .001. T-test analyses were also conducted to compare monolingual and bilingual performance within each block of the pre- and post-switch phases. As expected, the proportion of looking time to the correct location did not differ between groups in the pre-switch phase (Block 1: t(64) = .795, p = .429, d = .199; Block 2: t(65) = .758, p = .451, d = .188; Block 3: t(68) = .960, p = .341, d = .232), but contrary to expectation, this was also the case in all three blocks of the post-switch phase (Block 1: t(67) = 1.639, p = .106, d = .400; Block 2: t(68) = .432, p = .667, d = .105; Block 3: t(66) = 1.941, p = .057, d = .478). As seen in Figure 3 and Table 7, bilingual infants were at chance levels in all three blocks of the pre-switch phase (Block 1, t(28) = .273, p = .787; Block 2, t(26) = .963, p = .344, and Block 3, t(29) = .798, p = .436), and the first two blocks of the post-switch phase (Block 1, t(28) = -.288, p = .776, Block 2, t(29) = -.100, p = .921), and the proportion of looks to the correct location was only above chance in the third post-switch block (Block 3, t(28) = 3.219, p = .003). On the contrary, monolinguals were at chance in Block 1 pre-switch (Block 1, t(36) = 1.452, p = .155), but they were above chance in Blocks 2 and 3 (Block 2, t(39) = 2.529, p = .016; Block 3, t(37) = 2.474, p = .018). In the post-switch phase, monolinguals were below chance in Block 1, t(39) = -3.142, p = .003, and then again at chance in Blocks 2 and 3 (Block 2, t(39) = -.809, p = .423; Block 3, t(38) = .677, p = .502). These detailed analyses suggest that the source of the interaction lies in the differences in performance in the pre- and post-switch phases within each language group given that the between-group comparisons for each test block failed to reach statistical significance.

Discussion (Section 4). In paragraph 2 on p. 14, the sentences “Furthermore, in the post-switch phase of both the auditory and visual conditions, bilingual infants fixated the reward more than monolinguals. That is, despite their initial perseverance in anticipating the reward to appear in its pre-switch location, bilinguals were still more likely to fixate their gaze on the visual reward than monolinguals, even when it appeared in an unexpected location.” must be modified to “Despite this difference in anticipation patterns, bilinguals fixated the visual reward to the same extent as monolinguals in this condition, and in fact, they did so more than in the visual condition. That is, despite their initial perseverance in anticipating the reward to appear in its pre-switch location, bilinguals were still as fast as monolinguals in fixating their gaze on the visual reward, even when it appeared in an unexpected location.”

Re-analyses of the data resulted in different numeric values reported in the text, tables, and figures in the paper. In these cases, only the numeric values have changed, but the accompanying text has remained unaffected. The correct values reported in the text are presented in Table Corrigendum 1, and the corrected tables and figures are presented above. The Appendix, Online Supplementary Materials, and all the materials used for analyses (available at osf.io/k3e9z/) have been updated.

奖励窗口绩效（第 3.1 节，第 9 页）。当奖励出现在屏幕上高于机会水平（机会 = .5）时，婴儿总体上可能会注视奖励。线性混合效应模型的结果产生了阶段和块的显着主效应，以及显着的逐阶段和逐条件组交互。总体而言，婴儿在单语转换前阶段比转换后阶段更多地固定奖励，ß = -.008, SE = .001, CI [-.001, -.006], t = 6.019, p < .001，双语组，ß = -.004, SE = .002, CI [-.007, -.006], t = 2.354, p= .018，但双语者更有可能将奖励固定在听觉而不是视觉条件下，ß = .004, SE = .001, CI [.001, .007], t = 2.578, p = .010。

视觉状况（第 3.2.2 节，第 12 页）。与听觉条件类似，该模型产生了块和相位的主要影响。婴儿的表现也随着任务的进行而提高，婴儿在切换前比切换后更能固定正确的位置以期待奖励，ß = -.085, SE = .035, CI [ -.15，-.02]，t = 2.416，p = .016。Group by Phase 互动也很重要。计划的成对比较显示，单语和双语婴儿在转换前阶段的表现相似，ß = .075, SE = .050, CI [-.02, .17], t= 1.493, p = .136，但双语者在转换后阶段的表现优于单语者，ß = -.115, SE = .049, CI [-.21, -.02], t = 2.317, p = .021。在这项任务中，双语者在转换前后阶段的表现没有差异，ß = .010, SE = .053, CI [-.10, .11], t = 0.181, p = .856，但是与转换前相比，转换后单语者的表现显着降低，ß = -.179, SE = .046, CI [-.27, -.09], t = 3.909, p< .001。还进行了 T 检验分析，以比较转换前和转换后阶段的每个块内的单语和双语表现。正如预期的那样，在切换前阶段，各组之间寻找正确位置的时间比例没有差异（块 1：t (64) = .795，p = .429，d = .199；块 2：t ( 65) = .758, p = .451, d = .188; Block 3: t (68) = .960, p = .341, d = .232)，但出乎意料的是，这也是所有情况切换后阶段的三个块（块 1：t (67) = 1.639, p = .106, d= .400; 块 2：t (68) = .432，p = .667，d = .105；块 3：t (66) = 1.941，p = .057，d = .478)。如图 3 和表 7 所示，双语婴儿在转换前阶段的所有三个模块中都处于机会水平（模块 1，t (28) = .273，p = .787；模块 2，t (26) = .963, p = .344, 块 3, t (29) = .798, p = .436)，以及切换后阶段的前两个块 (块 1, t (28) = -.288, p = .776，第 2 块，t (29) = -.100，p= .921），并且在第三个转换后区块（区块 3， t (28) = 3.219，p = .003）中，寻找正确位置的比例仅高于机会。相反，单语者在 Block 1 pre-switch 中有机会 (Block 1, t (36) = 1.452, p = .155)，但他们在 Block 2 和 3 中的机会高于 (Block 2, t (39) = 2.529，p = .016；第 3 组，t (37) = 2.474，p = .018)。在转换后阶段，单语者在块 1 中的机会低于t (39) = -3.142, p = .003，然后在块 2 和 3 中再次出现机会 (块 2, t (39) = -.809 ,p = .423; 块 3，t (38) = .677，p = .502)。这些详细的分析表明，相互作用的根源在于每个语言组内切换前和切换后阶段的表现差异，因为每个测试块的组间比较未能达到统计显着性。

讨论（第 4 节）. 在第 2 页第 2 段中。14，句子“此外，在听觉和视觉条件的转换后阶段，双语婴儿比单语婴儿更关注奖励。也就是说，尽管他们最初坚持预期奖励会出现在转换前的位置，但双语者仍然比单语者更有可能将目光固定在视觉奖励上，即使它出现在一个意想不到的位置。” 必须修改为“尽管预期模式存在差异，但双语者在这种情况下对视觉奖励的固定程度与单语者相同，事实上，他们比在视觉条件下做得更多。也就是说，尽管他们最初坚持预期奖励会出现在其切换前的位置，但双语者在将目光固定在视觉奖励上时仍然与单语者一样快，

对数据的重新分析导致论文中的文本、表格和图形中报告了不同的数值。在这些情况下，只有数值发生了变化，但随附的文本不受影响。正文中报告的正确值列于表勘误表 1 中，更正的表格和数字列于上面。附录、在线补充材料以及用于分析的所有材料（可在 osf.io/k3e9z/ 上获得）已更新。