A study published in journal Science stated that ducklings are capable of identifying shapes and colors. However, some critics questioned the findings that it may have only stemmed from mere coincidence.
Ducklings Are Smarter Than We Thought, Says Study
Researchers Antone Martinho III and Alex Kacelnik from the University of Oxford published their findings on their duckling study showing they can identify “same” and “different” in shapes and colors. This headline-grabbing discovery caught the interest of many publications that it easily proliferated across social media.
What makes the ducklings even more exceptional, as per the study, is how they are not trained, unlike other animals that go through “extensive reinforcement training.” Knowing how the baby ducks have the mechanism called “imprinting” where they “imprint” on their mother as soon as they hatch to recognize her, the researchers wondered what it is on the ducklings’ brain that allows the functioning.
From the experiment, the baby ducks were shown objects (by color and shape) with similarities (for example, two yellow cubes) and differences (a cuboid and a sphere). After imprinting, two-thirds of the baby ducks were observed to choose the objects with the same characteristic they had imprinted on when shown with new objects.
Findings Are Mere Coincidence?
However, not everyone agrees. Two separate research papers raised questions about Martinho and Kacelnik’s findings after reanalyzing the data.
From the Leibniz Institute for Farm Animal Biology in Germany, Ian Langbein co-authored the study with journal Science’s Birger Puppe where they used other statistical and binomial tests. Langbein says the conclusion of the study only supported shapes and not colors since baby ducks can hatch at “any time of day and night.” With that, color cannot be a “salient stimulus” since brightness and color could differ during imprinting.
Another study conducted by Jean-Michel Hupé from the Université Toulouse argued with p-values, especially when they are below 0.05. His reanalysis demonstrated that little conclusion can be deduced from the duckling data. Hupé says using p-values is not just controversial, but it is just “wrong.”
Both studies, however, do not ask for a retraction. Hupé only wants his reanalysis to serve as “cautionary notice” while Langbein wants a “new interpretation.”
A representative from journal Science also reportedly spoke up that they are not retracting the findings from the paper. This is because the public exchanges allow transparency over debates and conclusions.
Martinho and Kacelnik also published their response to the criticisms they met. Martinho says they are doing “more experiments testing the same and related ideas” about the ducklings and so far, the results are “reassuringly strong.”