This video is adapted from 10.3390/biology14030297
A recent report suggested something striking: Japanese tits may use a wing-fluttering gesture as a symbolic “after you,” signaling a mate to enter the nest box first. Because symbolic gestures are often considered cognitively demanding, this claim has broad implications. This video reexamines the evidence by reanalyzing the shared dataset with bootstrapping and causal inference.
This video begins with bootstrapping, generating one thousand resampled datasets from the original observations. This lets the video ask a practical question: if the data is repeatedly “re-drawn” from the same source, how much do the key estimates move?
Two results stood out. First, the estimated proportion of “after you” gestures when a mate is present shows substantial spread, with a 95% bootstrap confidence interval from 18.9% to 37.8%. Second, male response timing varies as well. The bootstrap confidence interval for mean latency to enter the nest spans about 36.6 to 84.3 seconds. In other words, the gesture looks stable as a behavioral category, but its observed frequency and timing are not fixed.
This video also highlights context dependence. For females, wing fluttering is rare when the female is the first feeder, about 3.03%, but it is common when she is not the first feeder, about 95.83%, with an extremely small Fisher’s exact test p-value of 10 to the minus 13. In addition, a sex-by-wing-fluttering table yields Fisher’s p equal to 0.0002, and bootstrapping estimates the proportion of females displaying wing fluttering when a mate is present at about 42.09%, with a 95% interval from 29.82% to 54.39%.
Finally, this video asks whether wing fluttering is associated with quicker male entry. After merging and cleaning the datasets, 33 records remained. A comparison of response times between events with and without wing fluttering produced a t-statistic of minus 3.30 and p equal to 0.0048, consistent with faster male responses when wing fluttering occurs.
Taken together, this video’s reanalysis supports a communicative interpretation, while also emphasizing variability, small samples, and the need for careful replication.