Erjavec and Horne (2008) investigated the determinants of imitation in children. The general view, according to the authors, is that imitation is a function that is established early in human development and that makes possible the acquirement of important behavioral repertoires like verbal behavior that are exclusively human. There appears to be disagreement to if the propensity towards imitation in humans is innate or learned; however, the process of learning imitation is also of interest to researchers and several different models abound. The researchers used an accepted procedure of training infants in a baseline modeling phase of the same four behaviors with continuous reinforcement and then a 50% VR2 intermittent reinforcement schedule and then interspersing the baseline matching behaviors with unreinforced target behaviors. The dependent measure was the accuracy of the imitation of the target behaviors, whereas several independent variables were used including the age of the children (there were groups of three and four-year-olds), common or uncommon behaviors, and the number of trials. In essence the target behaviors are not reinforced (or are they?) and one explanation of the imitation of unreinforced behaviors is a sort of generalization as referred to by Pavlov but termed in these paradigms as parity. That is children generalize to like imitative reinforced behaviors, but because children produced common responses (trained) as mismatches to uncommon ones and did not improve in their accuracy with over subsequent trials Erjavec and Horne (2008) found no evidence for the parity explanation. A Skinnerian model was supported in that children's...

Children were noted to be biased towards mirroring the experimenters, that is that the child would imitate a left hand nose touch from the experimenter (who was directly across from them) with their right hand, and there was a slight bias for the use of the right hand.
Well, there are several major flaws in the procedure in terms of the behavioral training. The first question concerns the idea that the target behaviors were not reinforced. From a pure Skinnerian or Thorndike point-of-view they are being reinforced. What is being reinforced in the paradigm is the act of imitating the experimenter, not the particular behavior involved. For example, in the baseline training sessions (that were originally reinforced on a continuous basis) children received praise and a toy for imitating the action of the experimenter. According to Thorndike's law of effect behaviors followed by pleasant outcomes are repeated, those followed by unpleasant outcomes are not. So the children wanted to "play the game" and really were not concerned with mimicking the exact actions of the experimenter. Quite frankly the baseline behaviors are so general that what is being reinforced is playing and trying to imitate the experimenter, not doing exactly what the experimenter does. The target behaviors are much more specific than the baseline behaviors. According to Thorndike's law of exercise and law recency behaviors tend to be repeated and recently repeated behaviors attend to reoccur first. Thus, the children develop rapport with the experimenter and then are told that they are…