Learning To Blog & Blogging To Learn

Blog With Me?

Month: May, 2012

So far, we have…

So far, we have only looked at studies that used positive reinforcement to achieve behavior that is something positive for both the subject and those around him or her.  This week, we will look at how positive reinforcement can actually produce and increase behavior that may be harmful and even lead to one’s detriment.  In the article by Billings et al. we see a study on how “reinforcement can induce children to falsely implicate themselves in wrongdoing” (Billings et al. 125).  Their verbal behavior can be manipulated and shaped such that the accuracy of their statements is greatly reduced.  Depending on how an interview is conducted, an interviewer can induce false accusations and confessions of wrongdoings.  Additionally, these results are robust such that the children “continue to affirm their accusations even when challenged” later on (Billings et al. 126).  First let’s look at how such a phenomenon is achieved.  We will then look at why this phenomenon may occur.

Ninety-nine 5 to 9-year-olds are introduced to and are allowed to play with a colorful toy in their elementary school.  After a span of three days, the experimenters remove the toy and claim that someone has taken the toy, probably just to play with it.  They explain that they believe a student is responsible and that they need the children’s help to find out what has happened to it and who has taken it.  Each child is then taken separately to be interviewed and is questioned in a “warm, supportive way” (Billings et al. 128).  The experimenters manipulate how the children are interviewed such that the control group of children are asked straightforward suggestive questions while the reinforcement group children are approached with the same suggestive questions but with an addition of reinforcement after every question answered with a “yes.”  Reinforcement includes verbal responses such as “Thanks!” “Great!” “You’re being a real help.” 

There are five groups of interview questions:

1) Filler Question: these questions are expected to produce a “yes” answer.  (“Do you remember the Brain Warp toy?”)

2) Guilty Knowledge Questions: these questions ask for information the child cannot possess unless he or she is present when the toy is taken.  (“When the kids took the toy, did the toy make a loud noise?”)

3) Direct Witnessing Questions: these questions ask if the child has directly observed the toy as it is taken.  (“Did you see the kids take the toy?”)

4) confession Questions: these questions ask if the child has personally participated in the taking of the toy.  (“Did you help another kid take the toy?”)

5) Leading correct Questions: these questions ask the child for information that the child does in fact possess.  (“Did the toy have different colors?”)

The Guilty Knowledge, Direct Witnessing, and Confession questions are classified as the misleading questions that are potentially self-incriminating.  Results are calculated based on the percentage of times the children answer “yes,” either verbally or nonverbally, to one of these three categories of questions.  Billings et al. find that the simple reinforcement tactics can in fact “induce children to make false incriminating admissions against themselves” such that 52% falsely admit to guilty knowledge concerting the theft, 30% falsely admit to witnessing it, and 18% falsely confess to participating to the theft with their corresponding control group counterparts resulting in 36%, 10%, and 6% respectively (Billings et al. 133).

Billings et al. also find that “conversations with teachers afterwards [indicate] that the children generally [enjoy] being questioned, even though from a strictly legal viewpoint, they [implicate] themselves in an apparent theft” (Billings et al. 133).  Why then are children willing to admit to being part of or being responsible for something negative which may give rise to punishment?  Their sense of enjoyment may stem from the fact that this type of reinforcement causes them to feel like they are doing something praiseworthy and honorable at the immediate time frame. In this experiment, it is important to Billings et al. that everything is conducted such that the children have a positive frame of mind and lack any anxiety.  Although they are admitting to something that is condoned, they are more focused on their prevalent actions which are prompting praise.  Or it could be that due to their limited experience, the children may be failing to recognize that “by making false claims of guilty knowledge or direct witnessing, they [are] thereby implicating themselves in the theft” (Billings et al. 134).  Instead, they think that they are just helping and are not realizing the ramifications of their claims.  It could also be that they do understand the consequences and are paying attention to both, but merely outweigh the “negative utility of possible self-incrimination [with] the positive utility of receiving praise and approval from the interviewer” (Billings et al. 135).  They lack the ability to forego the immediate positive consequences to avoid the more remote negative consequences.

The reason that this is so important is because Billings et al. mention additional studies of adults and adolescents that find similar results.  Positive reinforcement and feedback can “induce adult eyewitnesses to make inaccurate statements or report false memories.  Older individuals will sometimes even make false self-accusations in response to social pressure and other influences” (Billings et al. 126-127).  Although the type of reinforcement and tactics used must differ slightly due to the age differences, this phenomenon still persists.  In regards to our legal systems, the consequences for an individual therefore can heavily depend on the interrogation and interview process and the level of experience each individual has.  How much conscious control do we actually have then of our outcomes and present situations?  In this study, “gentle tactics [. . .] are able to extract self-incrimination admissions from children and even several full confessions in a matter of a few minutes” (Billings et al. 136).  Emphasis must be placed on the gentleness of these techniques and the fast manner in which they work.  While being questioned, one may not realize how bad a situation they are getting into since the process may seem so demure and safe due to the positive aspect of the rewards.  In addition, one may think that nothing consequential can result from a questioning of only a few minutes.  This type of positive reinforcement therefore is an enemy in disguise that is harder to reveal than if being interrogated in a harsh, accusatory manner.  The limitation for this study, however, is that the crime is a theft-crime.  If the crime is instead something more serious, such as breaking the toy, will the results change? 

In closing, it is also important to mention of a case where this phenomenon actually happened.  In 1998, a 7-year-old and 8-year-old boy were charged for first-degree murder after admitting to killing an 11-year-old girl.  Luckily, evidence proved that they could not have possibly been the killers and the charges were dropped.  Investigators believed that the interviewers were responsible for the inaccuracies of the boys’ confessions.    The interviewers bought the children a Happy Meal and reminded them that “good boys tell the truth [. . .] held their hands, told them they were friends, and questioned them about the murder” (Billings et al. 126).  These things could all be seen as positive reinforcement and could have easily manipulated the boys’ statements.

 

Billings, F., Taylor, T., Burns, J., Corey, D., Garven, S., & Wood, J. (2006) Can reinforcement induce children to falsely incriminate themselves? American Psychology, 31, 125-139.

Advertisements

What happens when you introduce something to look forward to in a situation that is bleak and both a psychological and physiological struggle?  This can be the difference that allows for motivation to adhere to treatment regimens for those with medical difficulties.  In fact, the two articles we will discuss this week look at methods for improving adherence to treatment of chronic diseases in children via positive reinforcement.  The importance from these studies stem from the fact that poor adherence is a common problem which, in many cases, is the underlying cause of poor clinical outcomes. Therefore, this may “lead to misconceptions regarding the efficacy of the treatment,” sometimes creating what merely “appears to be a ‘treatment resistant’ disease” (Luersen et al. 5).  The treatment is then perceived to be the problem when, in actuality, the problem is the patient’s willingness to faithfully and correctly stick to the treatment.

Luersen et al. look at this problem of variable adherence and it’s importance as a factor in clinical outcomes.  They review other studies which implement sticker charts to instrumentally train children with chronic diseases to adhere to their respective treatment regimens.  Each time the child completes their specific treatment or reach specified intake levels of substances, the child receives a sticker.  This type of positive reinforcement via the application of a sticker is thought to do two things: (1) It provides the children with an immediate sense of gratification and (2) It helps remind the children (and parents in some cases) whether a treatment has been completed and when the next dose is due.  I will lightly touch upon the studies Luersen et al. focus on but please feel free to look into the specifics involved at http://onlinelibrary.wiley.com/doi/10.1111/j.1525-1470.2012.01741.x/pdf.  Following is an example of a sticker chart:

Image

In a study by Stark et al., 4-12 years old cystic fibrosis patients are rewarded with a sticker each time a calorie goal is met. When compared to control groups, results include greater increase in caloric intake, weight gain, and BMI which was maintained during follow-up testings.

In another study by Stark et al., 4-10 year old juvenile rheumatoid arthritis patients are rewarded in a stepwise fashion with one sticker per increase in calcium intake per meal and for meeting specified calcium goals.  When compared to control groups, results include an increased calcium adherence and greater body bone mass.  Patients 5-12 years old with inflammatory bowel disease also participate in this same type of sticker program which similarly results in an increase in calcium intake and a greater percentage of patients achieving their set calcium goals.

The study of Cass et al. involve tuberculosis patients who are 1-14 years of age.  The introduction of sticker charts in which they are rewarded one sticker for each daily tuberculosis medication ingested makes children 2.4 times more likely to complete their latent tuberculosis infection treatment.

Slifer et al. study 3-5 year old patients with breathing disorders and how the sticker charts increase nightly BiPAP (Bilevel positive airway pressure) use.

In a study by Penica et al., a 2 year old hemophiliac child is rewarded with a sticker for adherence during the IV infusion of clotting factors.  This results in both the decrease in negative behavior and increase in positive behavior during the infusion process.

Luersen et al. findings therefore support their claim that the sticker charts are effective at increasing adherence to therapy in children with chronic disease.  Moreover, the effect is robust, working across a wide variety of chronic pediatric diseases.  What’s important about this is that these interventions improved the clinical outcomes, which is the ultimate goal of the adherence modifying methods.  What is the reason for the apparent success of this approach?  One can speculate that introducing a reward program gives the patients something to look forward to and want to work for.  It is something positive in their already adverse and tiresome life.  Medical difficulties, especially chronic diseases, not only burden the body but also have significant psychological ramifications.  It is not normal for a child to have to go through the demanding treatments, to be restricted from everyday activities, to be constantly worrying about their health.  Therefore, giving them some kind of encouragement, something rewarding, something positive is extremely important.  The desire to adhere to their treatments is instilled through this instrumental positive reward.

To further understand this type of instrumental conditioning, specifically positive reinforcement, let us now go in depth into one of the studies which results in similar findings as the studies above. This four patient study by Magrab et al. include one 11-year old, two 13-year olds, and one 18 year-old, all of whom are in the pediatric hemodialysis unit receiving dialysis treatment 2-3 times a week.  A serious problem among all four subjects, which is actually common for many pediatric patients undergoing dialysis treatment due to renal failure, is adherence to strict dietary restrictions.  Failure to follow the prescribed diets results in “increase risk for fluid overload and congestive heart failure, hypertension, hyperkalemia, azotemia, and bone disease” (Magreb et al. 573).  Because few techniques are found to be successful at maintaining compliance to the dietary restrictions, Magrab et al. wish to test the success of a token economy on these type of patients.  Their token economy is a bit different than the conventional in that only reinforcement is delivered (never is anything taken away).  In order to receive a reinforcement of 2-3 points, weight gain between dialysis sessions must be equal to or less than 2 pounds.  Weight gain is used as a measure of fluid intake, and 2 pounds represents the largest amount of fluid safe for dialysis.  Eighteen points are equivalent to $2.00 for purchasing prizes or rewards.  What is unique about the specific reward system of this study when compared to the previous studies discussed is that it is a “specially for you” system.  The children are allowed to construct their own prize list and indicate specifically what rewards the want to work for (as long as it is not food).  In addition, a chart which records their respective point achievements is posted in a very visual location in the unit for further motivation and recognition.  Each week the child’s point count is gathered, and he or she is allowed to exchange the earned points for their predetermined prizes.

Thankfully, the results of Magrab et al. are promising.  Compared to the baseline data, a reduction in the amount of weight gained found is found between dialysis sessions such that the average baseline gain is 2.18 lbs while the average treatment gain is .97 lbs.  In addition, a substantial change in the percentage of sessions in which the patients exceed acceptable weight gains decreases from a baseline of 47% to 20% during treatment.  It is also noticed that any extreme weight fluctuation that occurs during baseline is reduced during treatment.

Again, it can be hypothesized that the reason why this type of reward system is successful is because many psychosocial adjustments must be made when one is ill and “severe restrictions or major changes in eating style may become further barriers to normal social activity” (Magrab et al. 573).  Therefore, many more negative things are introduced into the lives of these patients.  By introducing something positive, these children have something to look forward to, something to work for.  This essentially enhances the quality of their lives and provides a type of “life motivation” (Magrab et al. 578).  A reward system can also give the patients a sense of control over something in a situation they have little control over.  They gain a sense of power over their present states.  Moreover, the special thing about this case is that the children are able to choose exactly what they want their reward to be.  This then tweaks the system so that the motivation to complete the task, to adhere to their dietary restrictions, is greatest.  In the other studies, the value of stickers may be different from one patient to another.  However, in this study, the value of the reward is high for all patients.

 

Magrab, P. R. & Papadopoulou, Z. L. (1977) The effect of a token economy on dietary compliance for children on hemodialysis. Journal of Applied Behavior Analysis, 10(4), 573-578.

 

Luersen, K., Davis, S., Kaplan, S., Abel, T., Winchester, W., & Feldman, S. (2012) Sticker charts: A method for improving adherence to treatment of chronic diseases in children. Pediatric Dermatology, 1-6.

Last week, we looked at the patterns of reinforcement and discussed the experiment that proposed intermittent reinforcement as a better means of resistance to extinction than is continuous reinforcement.  This week, we will focus on the specific characteristic of the reinforcement and examine an experiment that takes advantage of embedded reinforcement.  Before explaining what is meant by embedded reinforcement, I must first make Hanley, Tiger, and Ingvarsson’s experiment familiar to you.

Haley et al. focus on free-play periods in the pre-school setting.  Free-play periods are characterized by “children initiated engagement [and] provide children with opportunities to choose from a variety of simultaneously available activities that are presumably consistent with their interests and abilities” (Hanley et al. 33). This type of free-play is commonly seen as an opportunity to develop social and academic skills.  The experiment of Hanley et al. emerges from the finding that “selection and engagement of materials in instructional, literacy, and science zones [are] consistently low compared to [that] in dramatic play, computers, blocks, manipulatives, games and art activities” (Hanley et al. 33).  Therefore, Hanley et al. propose two strategies to promote the selection of these less preferred activities:

1)   Satiation: By providing prolonged access to the preferred activities and keeping these activities constant (lacking any type of novelty), participation in these activities should decrease due to satiation or habituation.

2)   Embedded reinforcement: By adding attracting qualities to the locations of the less preferred activities, subjects will be lured to those activity zones and subsequently engage in those activities.

In the experimental set up, there were nine zones the children could choose to play in.

  • Dramatic play: Pretend play toys (e.g., dress-up clothes, doctor set, flower shop, barber shop)
  • Computer: Two computers with a variety of CD-ROM games (e.g., Clifford’s Counting, Jumpstart Kindergarten)
  • Blocks: Toys to occasion large motor movement (e.g., train sets, large blocks, bowling set, basketball)
  • Manipulatives: Small toys on table tops to occasion small motor movement (e.g., building blocks, animals, tinker toys, Lincoln Logs)
  • Games: Age-appropriate board games and large puzzles (e.g., Candyland, Memory, dominos)
  • Art: Open-ended art activities on table top (e.g., paint, crayons, Play-Doh)
  • Science: An open-ended activity for children to explore and use their senses (e.g., digging for dinosaurs in sand; pouring water through sieves)
  • Instructional zone: One-on-one direct instruction. Each child has individualized skills and relevant materials selected and stored in the area
  • Library: A variety of age-appropriate books

The last three zones are the zones that are initially least preferred during baseline.  During baseline, dramatic play, blocks, art, games, manipulatives, and science materials are rotated daily. Because of the large number of books in the library and the individualized nature of the instructional zone, the materials in these activities are rotated weekly. A wide variety of computer games are located by the computer; therefore, these materials do not rotate (Hanley et al. 35).  In the satiation phase, the materials in all zones other than the science, instructional zone, and library zone are kept constant.  In the embedded reinforcement phase, the following zones change:

Instructional zone: the chairs and cubbies of the instructional zone are redecorated with popular children’s cartoon characters;  when possible, a teacher sits in the instructional zone prior to children selecting the instructional zone; small trinkets are placed intermittently in each child’s bin of individualized instructional materials, which are available to children who select and sit in the instructional zone (Hanley, et al. 37).

Library area: the table and chairs are replaced with four plush pillows and a carpet in the library; a book of the week is selected and displayed with thematically related toys; when possible, a teacher sat in the library prior to children selecting the area (Hanley, et al. 37).

Science: Science-related activities that are thought to provide more reinforcement are arranged in the science area. To increase the likelihood that children will select the science area, teachers present each new science activity during group instruction the day prior to its inclusion in the science area (Hanley, et al. 37).

Hanley et al. conduct the experiment by first holding the baseline phase, then holding the satiation phase, baseline phase, and finally the embedded reinforcement phase.  The results seem to indicate that an embedded reinforcement type of procedure produces more success overall.  After the satiation phase, there is a slow decline in allocation to dramatic play and blocks but no significant change in any of the other zones (other than the three of interest: instructional, science and library).  The instructional and science zones feel the indirect effects such that there is an increase in preference for those; however, there was no change for the library zone.  After the embedded reinforcement phase, there is an immediate increase in attendance all three zones of particular interest, the library, science and instructional zones.  In addition, these effects are sustained and present during the follow up testings.

These results are interesting yet, I question whether the seemingly success of the embedded reinforcement is due to the fact that satiation occurs prior.  Maybe it is not just the embedded reinforcement that is causing the desired results.  Maybe there is an effect of both together (as long as they are in near temporal vicinity to each other) that causes the increase in the instructional, library, and science zones.  It could be that the reason why we interpret it to be due to the embedded reinforcement is merely because embedded reinforcement was the second strategy performed.  To solve for this, further research can switch the order such that the embedded reinforcement is before the satiation.  If the same effect is found such that the embedded reinforcement causes stronger, more sustaining results, then it is in actuality due to embedded reinforcement.  The two strategies could also be tested separately using a between subject design.  All could be combined to see if the effects are due to one or the other, or to see if combining the two creates even better results.

Another thing to discuss is this method of embedded reinforcement.  To my understanding, I interpret from Hanley et al. article that this type of reinforcement is just to attract the children over to the library, instructional, and science zones.  The children do not necessarily have to partake in the activities to be rewarded.  Therefore, this method depends on the probability that if they are in those zones, they will interact with the respective activities because they are nearby.  The embedded reinforcers are there just to call over the children and get them to pay more attention to the zones they would normally prefer less.  They are not rewarded for taking part in those zones; they are rewarded for being in those zones.  The experiment of Hanley et al. show that apparently, this is successful.  It could be then suggested that reinforcement is a very powerful tool since, even though the reinforcement is not directly on the behavior of interest, it still produces the desired, indirect effects on that behavior.

Hanley, G. P., Tiger, J. H., & Ingvarsson, E. T. (2009) Influencing preschoolers’ free-play activity preferences: an evaluation of satiation and embedded reinforcement. Journal of Applied Behavior Analysis, 42(1), 33-41.

Thus we begin the second half of the quarter…

We have already discussed the success of instrumental conditioning, specifically, that of reinforcement training.  To go one step deeper, we will now focus on more detailed aspects of this strategy.  We will look at how exactly reinforcement is administered.  An experimenter by the name of Bijou focuses on the specifics of reinforcement and hypothesizes that “intermittent reinforcement (whether fixed or irregular in pattern) markedly increases resistance to extinction as compared to continuous reinforcement” (Bijou, 47).  To discuss this, it is helpful to define the most common types of reinforcement patterns as quoted in the article:

1. Continuous reinforcement: In the experimentally defined situation, a response is reinforced on each occasion of its occurrence.

2. Intermittent reinforcement: A reinforced occurrence of a response is preceded or succeeded on at least one occasion by an unreinforced occurrence of the response.  No differentiation is made among the terms descriptive of this procedure; namely, intermittent reinforcement, partial reinforcement and periodic conditioning.

a) Interval intermittent reinforcement: The pattern of a reinforcement is controlled by temporal events in the external environment.

b) Ratio intermittent reinforcement: The pattern of a reinforcement is dependent on the subject’s behavior and follows a specific ratio.

c) Fixed and Variable patterns of intermittent reinforcement: The relationship between the reinforced and nonreinforced is either fixed or variable.  Both interval and ratio may be either fixed or variable in pattern.

Bijou’s experiment focuses on the following question: “For a given number of reinforcements, is there a difference in the extinctive behavior of two groups of preschool children when the training of one group is on a continuous reinforcement pattern, and the training of the other group is on a variable intermittent schedule with reinforcement following 20% of the responses?” (Bijou, 48).  To answer this, Bijou uses a box with two holes, one above the other such that the top is the input hole and the bottom is the output hole.  The subjects are allowed to place a rubber ball in the top hole which is the action that is rewarded.  A motor driven machine is also utilized to dispense trinkets as rewards.    The experimenters explain to the subjects how the box works and allow the children to play with the apparatus.  One group of children, group A, are rewarded with trinket reinforcements six times in a consecutive order.  The second group of children, group B, are rewarded with trinket reinforcements six times such that they are received over 30 responses, specifically on trials 1, 6, 13, 17, 23, and 30.  For both group A and group B, the extinction period followed immediately after such that the children received no reinforcement for three and a half minutes regardless of their behavior.

From the data collected it is clear that the rate of extinction is greater for the group that was continuously reinforced (group A: 100%) than for the group that was reinforced intermittently (group B: 20%).  The mean number of responses for the extinction period is 15.3 and 22.0 for group A and group B respectively.

Bijou goes on to do a second experiment where now, the trinket dispenser is accentuated by a buzzer.  Everything else is identical to the first experiment.  The same general trend occurs where the continuously reinforced group shows a greater extinction rate than the intermittently reinforced group.  The mean number of responses for the extinction period is 13.0 and 26.2 for group A and group B respectively.

From Bijou’s experiments, we are provided with two interesting points.  First, “for a given number of reinforcements, a variable ratio intermittent distribution is associated with more resistance to extinction than a continuous schedule” (Bijou, 52).  This is interesting because one would intuitively think that a continuous schedule would better reinforce a behavior than would an intermittent schedule.  An increase in contingency usually is related to greater learning.  Therefore, one might assume the robustness and resiliency of such a strongly learned behavior would prevail and resist extinction.  However, this is not the case.  What could account for this?

Hypothesis 1) Because an intermittent type of reinforcement does not reward the subject on every trial, the subject could just assume it is just one of those trials that he or she will not rewarded.  They realize that, to be rewarded, they have to endure these trials with no reward and therefore will not be able to tell that they’ve entered an extinction phase or that anything has changed.  However, if the continuously reinforced subjects are not rewarded, this is not “normal,” and they realize that something has changed.  The “surprise” factor is different.

Hypothesis 2) The intermittent type of reinforcement already establishes a moderate amount of frustration on the subjects.  They are a bit upset that they do not receive rewards all the time, but this negative feeling is minor.  The continuous type of reinforcement establishes no frustration on the other group of subjects since they are always rewarded.  Although once no reward is administered, both group A and group B will be greatly frustrated, the difference between the initial levels of frustration and the final levels of frustration will be different for group A and for group B.  Group A will feel a greater change in frustration while group B, already frustrated a bit, will feel a smaller change in frustration.

The second point is that the difference in the mean amount of responses between group A and group B are higher in the second experiment than it is in the first experiment (Bijou, 52).  This means that the increased distinctiveness of the auditory stimulus serves as a stronger conditioned reinforcer.  The buzz increases the saliency of the reward, and because the difference between the means will be greater.

As we conclude this post, there is one thing I would like to mention.  This study held the number of reward trials constant such that both group A and group B received 6 rewards.  However, because group B was rewarded intermittently, group B participated in the experiment for a longer amount of time and for more trials (group A had only 6 trials, all of which were rewarded, and group B had 30 trials, of which 6 were rewarded).  This difference could account for the results found as opposed to the type of reinforcement pattern.  Therefore, to investigate this, future research can be done such that the amount of total trials, the sum of the reinforced and nonreinforced, is held constant.  The duration of the experiment is the same for both groups.  The comparison of this data to the data found in Bijou’s article can then explain this phenomena further and more accurately.

 

Bijou, S. W. (1957). Patterns of reinforcement and resistance to extinction in young children. Child Development, 28(1), 47-54.

Last week, we focused on the parent, specifically the mother, as a social reinforcer for certain behaviors in children.  Keeping in line with an external individual as the means and focus to ultimately alter the child, we will now look at the child’s peers as the social reinforcers.  However, the matters we talk about in this post have prominent differences and are of quite a different nature than that of the last post.  One obvious difference is the age gap between the reinforcer and the child.  The child’s peers are usually the same or around the same age as the child.  Another difference is the social role peers play.  While the mother’s role is to raise the child, the peers role is to merely provide companionship, support, and be a social equal.  The relationship between the child and peer greatly contrasts the relationship between the child and parent.  Therefore, how the child views the reinforcer is much different if the reinforcer is a peer or if the reinforcer is a parent.  The time spent with the reinforcer also differs.  As children grow up and begin to go to school or become involved in extracurricular activities, they sometimes begin to spend more time with their peers than with their parents.  All these are important differences to keep in mind while this post unfolds.

Patterson and Anderson’s article reveals how peers can serve as agents dispensing social reinforcers.  They hypothesize that “after extended experience with a peer group, the child responsive to social reinforcers from the peer group would be expected to show high frequency of behaviors valued by this group” because those are the behaviors that are most likely to elicit social reinforcers from the peer (Patterson & Anderson, 952).  Therefore, their first step is to prove that peers can in fact serve as social reinforcers.

To do so, Patterson and Anderson show that the child’s peer can be the social reinforcer to condition a simple motor response.  They use a box with two identical holes on the top into which the child can drop marbles.  The child is instructed to pick up the marbles one at a time and drop them into any hole they want.  They can drop as many marbles they want and in any order.  The frequency of response into either hole is recorded in respect to time.  After 100 marbles are dropped into the box, the experimenters establish a minimally stable baseline estimate of choice behavior; in other words, they determine which hole is preferred by the subject.  The subject then sits with the box right in front of him or her while facing one of their peers.  The peer serves as the reinforcing agent and is instructed by the experimenter through an earpiece to read the words “good,” “yes,” “great,” “ok,” “fine,” or “very good” when the subject places a marble into the least preferred hole.  Patterson and Anderson obtain the measure of preference change by computing X/(A+B) (where X is the frequency of the least preferred hole, A is the frequency of one of the holes, and B is the frequency of the other hole) before and after the reinforcement from the peer takes place.  The difference between these two values then “provides a measure of the magnitude of shift in preference” (Patterson & Anderson, 954).  Their data, which shows a large shift in preference, supports their hypothesis that a peer can act as a social reinforcer such that “the peer is clearly effective in changing the behavior of the subject” (Patterson & Anderson, 955).

Patterson and Anderson then extrapolate this phenomenon to say that if “the child is responsive to social reinforcers delivered by the peer group,” then “the child will show an acquisition of the behaviors valued by the peer group” (Patterson & Anderson, 958).  The child will be rewarded depending on the desirability of his or her behaviors and thus, the more desirable the trait, the more frequent this behavior will become.

Although this is a very interesting and exciting hypothesis, I would like to point out some limitations.  They are not to depreciate Patterson and Anderson’s finding or discredit their argument.  They are instead just some things to keep in mind, something future researchers can address.  Firstly, testing only peers as reinforcers may not be enough to provide sufficient data.  They do not, in addition and for comparison, use a parent, teacher, stranger, or other type of non-peer as an agent to disperse the same reinforcement.  Therefore, it is difficult to claim that the success of the reinforcement is due to the reinforcer being a peer specifically.  It could merely be because someone, anyone, is reinforcing the child’s behavior.  In regards to the apparatus that is used, there might be an inaccurate sense of hole bias.  If the child wishes to place the marbles in the holes following a certain pattern (for example: right, right, left, right, right, left…) this can give a sense that he or she prefers one hole (the right hole) when in actuality, this is not the case.  The child merely prefers the pattern.  In addition, the type of reinforcement received in the environmental setting is a bit unrealistic to the reinforcement the child would get in the natural setting.  The peer reinforcer only states mundane words (“good,” “yes,” “great,” “ok,” “fine,” or “very good”) in the experiment, while in a real life setting, a peer reinforcer might have much more active response.  This could include more colorful and descriptive verbal approvals, expressive body gestures, and more personal interaction.

—————UPDATE—————

Lastly, it is important to always have a control group in which to rule out any spontaneous preference of the other hole.  For example, a child may seem to respond to the peer reinforcement when in actuality, the child may just feel like putting the marble in the hole that so happens to be the least preferred.

Patterson, G. R., & Anderson, D. (1964). Peers as social reinforcers. Child Development, 35(3), 951-960.