Can an app really tell us why a baby is crying? Here is the pseudoscience behind these "baby translators"
Remember the translator for babies invented by the half-brother of Homer Simpsons, Herb Powell? In the famous episode, Herb is still trying to return to business after Homer destroyed his car company, and while he is a guest of the terrible half-brother he has the idea for theinvention of the century: a device that translates the sounds of newborns into expressions understandable by adults. Thanks to the adorable guinea pig Maggie, and at a (dutiful) loan from Homer, Herb Powell manages to build and market a device dreamed of by all the parents of the world.
When it came out this episode was the 1992, and baby translators could only exist in cartoons. Today things are still the same way, but if there is demand, why shouldn't the market satisfy it? Since when is the plausibility of a product a limit? Just think of homeopathy or, to stay on topic, amber necklaces that should alleviate teething pains in newborns (sic).
Today then entrepreneur barkers have a huge advantage over Herb Powell: everything you need for lighten the anxious parents one app, because the manufacturers of phones and tablets think about designing and building the hardware. By strokes of Press releases promptly re-launched (also in Italy), the software that decrypts the screams of newborns is passed off as state-of-the-art technology, the result of serious studies certified by respected scientists, but nothing could be further from the truth.
The app of the momentIn a recent post on Science Based Medicine the pediatrician Clay Jones demolished Infant Cries Translator, one of the most famous translators of the moment. The app presented as the product of researchers led by the engineer Chuan-Yu Chang, professor at the National Yunlin University of Science and Technology to taiwan, and by the pediatricianDr. Chen Si-da. In the press releases, as in a Reuters video, it is explained that the researchers built a database of 200,000 records of the crying of 100 babies, in which they were found subtle acoustic variations that make translation possible.
But how did you go from recording of tears to the program not given to know. Dr. Jones speculates that a must have happened miracle, because it doesn't exist no publication in this regard, or even any explanation on the principles of operation. Later a group of users would try the app, and based on this that manufacturers can state in the press releases that the translator is accurate up to 92% of the time for babies within two weeks of life. Subsequently the accuracy drops but, the app remains theoretically usable up to 4 months with an accuracy of 77%.
The study in SpanishBut Infant Cries Translator certainly not the first product of its kind to be marketed trying to exploit the respectability of science. Jones even found a clinical trial on the effectiveness of Cry Translator, another translator that exists since 2009 (and now also available as an app). The study was published in the journal Pediatrician Rural y Extrahospitalaria, which as you can guess does not foresee peer review of the contributions invited, and infati Jones defines comic the method used to validate the device. The study not only he was not blind, but the translator passed the test if the suggested intervention (e.g. cradling the infant) was successful. The problem that every single intervention included a form of interaction with the child, so the 96% accuracy result tells us just what we already knew: if a child cries for any comfort action, most of the time, he tends to calm down quickly. In all likelihood also the effectiveness heralded for the translator Made in Taiwan can be attributed to a confirmation bias of this type.
Why does a baby cry?A parent does always worry for their children, and companies know that to alleviate his anxiety that parent willing to spend small and large amounts. The problem that a specimen of Homo sapiens of a few days it doesn't have much to communicate who cares for it, and you can be sure that if a elementary language adults wouldn't need a program to interpret it. Do we really think that for millennia children have launched specific signals into the void until the arrival of the suitable technology? As Dr. Jones explains, newborn babies cry when they are hungry or feel uncomfortable, but sometimes they cry without identifying a reason.
The point that most of the time we don't really know why babies cry, and it doesn't matter how much experience or training an observer has, or how large the database of researchers working on it, or 'subtle acoustic variations' they discover.