Taking Inspiration from the Methods

Occasionally, while I'm doing background reading for my research project, I come across some interesting papers. Some of them are not directly relevant to my project so I keep them in a special folder on my laptop to make up for some casual reading later on. Last week I printed a couple of them, including one by Hamlin et al. published in Nature three years ago.

The paper reported that preverbal infants (aged six and ten months of age) assess the social behavior of individuals to judge whether said individuals are appealing or aversive. What's new or different about this? Well, it struck me as something we all think adults do, not infants. The methods used by the researchers were particularly interesting. As someone who will get into scientific research soon, I have everything to learn when it comes to designing an experiment — and scrutinizing the methods sections of scientific papers gets me accustomed to how scientific research is conducted.

Hamlin et al. were both astute and simple in their experimental design. To investigate the infants' intuitions about social interactions, the experiments relied on a choice paradigm in which infants reach to their preferred choices when presented with different things.

The first experiment made use of three simple toys as three distinct animated characters with a hill in the background. The circle character was the climber, the triangle one was the helper, and the square character was the hinderer. To the right is an approximate representation of the three characters used in the first experiment.

In one scenario, the triangle would help the circle climb up the hill by pushing it upward. In a second scenario, the square would hinder the circle's effort to climb the hill by pushing it downward. After witnessing these two scenarios, the infants were presented with the triangle and the square toys and allowed to choose one of the two (the choice paradigm). The infants' choices would then show which one they preferred of the two. A very simple and effective experiment it proved to be as the results were unequivocal. All 12 six-month-old infants and 14/16 ten-month-old infants chose the helping triangle. This choice, the researchers deduced, showed that infants had developed distinct impressions of the two characters on the basis of their actions toward the climber.

But what are the infants really choosing? Are they choosing the actions of the shapes, or do they just prefer triangles over squares? If the infants like the triangular shape or an upward motion more than a square shape or a downward motion, it would mean that they were not evaluating the objects based solely on their social behaviors.

To parse this out, a second "control" experiment was conducted with a different set of infants of similar ages. The experiment was identical to the first one, save one shrewd difference. And this is a real beauty. This time the pushed circle character was not a character at all but rather an inanimate object without a pair of eyes. Removing the eyes meant that the social notions of aiding and hindering would no longer apply in the two scenarios. Instead, the scenarios would now be viewed as a triangle character pushing a circle object up a hill and a square character pushing a circle down a hill respectively. To the right is an approximate representation of the two characters and the one object used in the second experiment. (Note that the circle has no eyes in this experiment, making it an inanimate object.)

If the choices of the infants in the first experiment were driven by factors other than the characters' social behaviors, the result of this second experiment would match that of the first experiment. This was not the case. Only four out of the 12 six-month-old infants and six out of the 12 ten-month-old infants chose the triangle this time. In essence, a more or less equal number of infants chose the triangle over the square.

So what did this all mean? These infant choices can be explained in three different ways: 1. they form good impressions of individuals seen helping another, 2. they form bad impressions of individuals seen hindering another, or 3. they do both of those. Hamlin et al. were not content with having to settle with those three possibilities. They thus performed a third experiment to pinpoint to the exact evaluation process.

Again, a different set of infants was used for the experiment. This time, the circle character (with its eyes) made its comeback, along with the triangle and square characters. To the right is an approximate representation of the three characters used in the third experiment. And yes, circle character is back in business!

Half of the infants were shown the helping scenario of the first experiment while the other half was shown the hindering scenario of the first experiment. Then they were each shown a different scenario wherein the circle character remained inactive or neutral. Instead of attempting to climb up the hill, it merely satisfied itself with remaining motionless at the foot of the hill. The infants shown the helping scenario then saw the triangle make the same uphill motion without pushing the circle character. The infants shown the hindering scenario instead saw the square character make the same downhill motion as they had previously seen though without any interactions with the circle character either. To simplify, half of the infants witnessed a helping triangle character and a neutral circle character while the other half had witnessed a hindering square character and a neutral circle character. When asked to choose between the helper and the neutral, 7/8 of both the six- and ten-month-old infants opted for the helper. The same proportion chose the neutral over the hinderer in the second group of infants. So, the order of preference from most popular to least popular is: helper, neutral, hinderer.

These, in essence, simple experiments showed for the first time that young infants' social preferences are influenced by others' behavior toward unrelated third parties. This ability for social evaluation so early in infancy, according to the researchers, may even be the foundation for moral cognition (e.g., the more complex notions of right and wrong). This is a pretty major statement. But when you consider that the infant subjects had no history with the characters nor did they themselves experience any consequences of these characters' actions, it becomes clear that the infants made judgments about the value of a social act.

Admittedly, Hamlin et al. are not in the same field of research as I will be next year, but the design of their experiments still left me in awe. I am very much looking forward to designing my own experiments in the future and, hopefully, they will be as elegant, simple, and thoughtful as those of Hamlin et al. In the meantime, it's the learning curve for me.

Reference:

Hamlin, J.K., Wynn, K., & Bloom, P. Social evaluation by preverbal infants. Nature 450, 557-560 (2007).