Alfred J Nigl, Ph.D. and Dean Grey
Skylab is one of the world’s most robust, white-labeled social media engagement platforms and was created to leverage the principles of the science of engagement. Its scientific foundation seamlessly integrates and operationalizes the research findings from such disciplines as Social-Cognitive learning, self-reinforcement and psychological gamification.
Skylab serves as a model for a well-founded, scientifically-based, social media gamification platform. Many authors have mentioned the science of engagement in the context of other social media platforms but few have defined the term and outside of vague references to motivation; very few scientific discussions and explanations of this ubiquitous term can be found online.
A specific search for “the science of engagement” will not yield much of substance. This paper is an attempt to provide a scientific and theoretical context for “the science of engagement” by referencing Skylab’s unique approach to integrating behavioral science into its highly-effective app and platform.
Many businesses today are focused on developing websites and/or apps which are designed to create a psycho-social platform for stimulating and/or optimizing engagement with their users, customers or subscribers. Engagement is a core principle of many internet-based companies. In fact, the greater the level of engagement, the more valuable a company becomes to both advertisers and investors. The rapid growth of social media advertising revenue is a testament to the economic value of engagement.
There is a considerable body of evidence documenting the rapid increase in the spending on social advertising. A fourth-quarter report from data science and media technology provider 4C Insights found a 43 percent quarter-over-quarter increase and a 65 percent year-over-year gain in paid media spend. Spending on Facebook grew even more, with a 74 percent Year over Year increase.
On Instagram, which saw user numbers rise to more than 600 million, advertising spend soared 138 percent from 2015; clearly, advertisers see the value of engaging their audiences as they transition to more platforms and networks.
The problem is that there are very few, if any, formal scientific studies on “engagement”; it is one of those ubiquitous terms (gamification is another one) that is used frequently today but when pressed to explain how it works and, especially, why it works, most people would be hard-pressed to give a logically-coherent explanation. There is a good reason for this, it is because currently there is no formal “Science of Engagement”, nothing tangible that someone can fall back on or use as a reference when trying to explain the term. But all is not lost. There is a legitimate scientific tradition which has existed since 1905 with a robust body of peer-reviewed scientific publications and scolarly textbooks that is a logical choice to serve as the scientific foundation of “engagement”.
Section I- Psycho-social Drivers of Engagement
At its most basic level, Engagement is a learned behavior, people become engaged using an app or interacting with others on a web site because these engagement behaviors are reinforced. That is, they learn to be engaged or to interact socially. In fact, many of the engagement behaviors seen today are not natural for many individuals (although there seem to be definite generational differences in this regard; e.g., many Millennials exhibit strong tendencies toward engagement which are not generally seen in older generations).Without strong reinforcers present, many people would keep to themselves and not actively engage with strangers. In fact the fear of social embarrassment (or criticism) is a strong driver of non-social (i.e., non-engaged) behavior even among Millennials. Obviously, anonymity helps but eventually actively engaged individuals, especially those who are identified as influencers must reveal themselves and throw off the mantle of anonymity to be an effective engagement influencer. Why would a person do that, the answer may lie in understanding how social-cognitive learning works.
The science of behavioral psychology has produced thousands of studies on the consequences of behavior reinforcement over the past 75 years. Behavior is said to be reinforced when the behavior is strengthened (e.g., occurs more frequently, occurs consistently and/or lasts for an extended period of time (i.e., it is not easily extinguished). Behavior reinforcement is linked to both learning and memory in humans and animals and it is the driving force behind the explosive adoption of social media channels by consumers and businesses alike. People are attracted to websites and apps which allow them to communicate and interact with people who possess similar likes and dislikes and a site or an app which satisfies a need or a want. These motivational factors are also an important component for learning to take place.
But a number of studies have shown that motivation is not the only factor that is important in learning a new action or behavior. As a reaction against Freudian doctrine of unconscious needs and drives, behavioral psychologists such as Thorndyke and Watson, developed the concept of Reinforcement to explain how a new behavior is learned by animals and humans.
Section II – Reinforcement Models of Learning
There are two main types of Reinforcement or conditioned learning that have been studied by behavioral psychologists:
Classical or Pavlovian Conditioning
Operant or instrumental Conditioning
Classical and operant conditioning share many of the same basic principles and procedures. For example, Kimble (1961) has pointed out that the basic principles of acquisition, extinction, spontaneous recovery, and stimulus generalization are common to both types of learning. There are several differences, however, between classical and operant conditioning. Although a basic feature of operant conditioning is reinforcement, classical conditioning relies more on association between stimuli and responses. A second distinction is that much of operant conditioning is based on voluntary behavior, while classical conditioning often involves involuntary reflexive behavior. However, research has shown that these distinctions are not as strong as they once were believed to be.
For example, Neal Miller (1978) has demonstrated that involuntary responses, such as heart rate, can be modified through operant conditioning techniques (which is one of the key foundations of the biofeedback technique (Nigl,1978;Fischer-Williams, Nigl and Sovine, 1976). It now appears that classical conditioning does involve reinforcement. And many classical conditioning situations also involve operant behavior. For example, let’s assume that a research subject (who we will refer to as Carol) was conditioned to fear large, aggressive dog breeds like German Shepherds or Pit Bulls. In the experimental environment, Carol would first learn to associate this type of dog with a sudden loud noise through classical conditioning of her alarm response which involves her sympathetic nervous system. Then the subsequent presentations of the dog would produce a fear reaction and Tina would learn to escape from the aversive stimulus through operant conditioning (negative reinforcement). This is sometimes called the two-factor theory of avoidance conditioning (Mowrer & Lamoreaux, 1942).
Since most, if not all, social media engagement scenarios involve voluntary behavior , operant conditioning is the primary mechanism governing this behavior.
Operant Conditioning and Cognitive Learning
The father of operant conditioning theory and practice is B. F. Skinner whose seminal research expanded the earlier work of Watson and Thorndyke and, in particular, his work was based on Thorndike’s (1905) law of effect. Skinner introduced a new term into the Law of Effect – Reinforcement.
Behavior which is reinforced tends to be repeated (i.e., strengthened); behavior which is not reinforced tends to die out-or be extinguished (i.e., weakened).The work of Skinner was rooted in a view that classical conditioning was far too simplistic to be a complete explanation of complex human behavior. He believed that the best way to understand behavior is to look at the causes of an action and its consequences. He called this approach operant conditioning. Operant Conditioning deals with operants – intentional actions that have an effect on the surrounding environment. Skinner set out to identify the processes which made certain operant behaviors more or less likely to occur.
Behavior modification is a set of techniques based on operant conditioning (Skinner, 1938, 1953). The main principle comprises changing environmental events that are related to a person’s behavior. For example, the reinforcement of desired behaviors and ignoring or punishing undesired ones.
There are different types of positive reinforcements. Primary reinforcement is when a reward strengths a behavior by itself. Secondary reinforcement is when something strengthens a behavior because it leads to a primary reinforcer.
Perhaps the most important finding to emerge from Skinner’s research has to do with the importance of schedules of reinforcement. The rate at which reinforcements are delivered can have a very positive or a very negative effect on the behavior being reinforced. The table below shows the different schedules of reinforcement that have been studied experimentally on animals as well as humans.TypesDescriptionEffect on Response Rate*Effect on Extinction**1ContinuousUser or subject is reinforced every time behavior is observedSlow rate of responseHigh rate of extinction once reinforcement stops2Fixed RatioReinforcement comes only after a certain number of responses occursFast rate of responseMedium rate of extinction3Fixed IntervalReinforcement comes after a certain time level has passedMedium rate of responseMedium rate of extinction4Variable RatioReinforcement occurs at an unpredictable rateVery fast rate of responseVery slow rate of extinction5Variable IntervalReinforcement given after an unpredictable time period has passedFast rate of responseSlow rate of extinction
How often the target behavior occurs (rate of response or rate of learning)
* How fast the target behavior fades away or disappears
One of the most well-known examples of behavior modification applications is the token economy which has been applied in hospital settings and special education facilities. Token economy is a system in which targeted behaviors are reinforced with tokens (secondary reinforcers) and later exchanged for rewards (primary reinforcers).
Tokens can be in the form of badges, buttons or virtual reward symbols, while the rewards can range anywhere from promotional offers or coupons to various forms of social or community recognition or privileges. Nigl (1978) created a token economy system that proved to be quite effective in changing behavior among profoundly and severely retarded adults and improved their social behavior as well as their personal hygiene and actually reduced the need to sedate these patients to control behavioral outbursts.
Going Beyond Traditional Operant Conditioning Reinforcement Systems
Many social engagement websites and apps have established quasi-token economies to reward positive engagement behaviors.
However, the truly successful companies (as measured by their average user engagement levels) have gone beyond a simplistic operant conditioning model to a more complex system involving a cognitive-based self-reinforcement model. One excellent example is Skylab. For example, recent statistical analysis of user behavior across Skylab’s multiple websites or “planets” shows that the average % of monthly active users is 46% which is considerably higher than stats reported by Youtube and Twitter and is more than 50% higher than other popular sites such as Instagram.Planetactive userstotal usersMAU %1Hybrid Central 30534688%2Powur PBC45352986%3TEM12214186%4Elevated10112084%5Cheer2371485748%6Ultimate Game29863147%7Gravity7315447%8Mind Movies4610245%9Allysian6972,40329%10Trilogy354368%11Totals4,5019,71946%
The graph below shows Skylab MAU % (Monthly Active Users) compared to other more established social media platforms.
Section III-Skylab’s Value Reinforcement System (VRS)
A. Social Learning and Cognitive Learning Theories
Social learning theory is a theory of learning and social behavior which proposes that new behaviors can be acquired by observing and imitating others. It states that learning is a cognitive process that takes place in a social context and can occur purely through observation or direct instruction, even in the absence of motor reproduction or direct reinforcement. In addition to the observation of behavior, learning also occurs through the observation of rewards and punishments, a process known as vicarious reinforcement. When a particular behavior is rewarded regularly, it will most likely persist; conversely, if a particular behavior is constantly punished, it will most likely desist. The theory expands on traditional behavioral theories, in which behavior is governed solely by reinforcements, by placing emphasis on the important roles of various internal processes in the learning individual
Social learning theory integrated behavioral and cognitive theories of learning in order to provide a comprehensive model that could account for the wide range of learning experiences that occur in the real world. As initially outlined by Bandura and Walters in 1963 and further detailed in 1977, key tenets of social learning theory are as follows:
Learning is not purely behavioral; rather, it is a cognitive process that takes place in a social context.
Learning can occur by observing a behavior and by observing the consequences of the behavior (vicarious reinforcement).
Learning involves observation, extraction of information from those observations, and making decisions about the performance of the behavior (observational learning or modeling). Thus, learning can occur without an observable change in behavior.
Reinforcement plays a role in learning but is not entirely responsible for learning.
The learner is not a passive recipient of information. Cognition, environment, and behavior all mutually influence each other (the principle of reciprocal determinism).
An important factor in social learning theory is the concept of reciprocal determinism. This notion states that just as an individual’s behavior is influenced by the environment, the environment is also influenced by the individual’s behavior. In other words, a person’s behavior, environment, and personal qualities all reciprocally influence each other.
For example, an engaged participant (a young girl age 7) in Skylab Cheer Life is recognized for giving her mother and other family members compliments at a very high daily rate (DUA); other girls who become aware of her activity from the Leaderboard postings begin to emulate her behavior so they can earn similar recognition and rewards.
This is an example of reciprocal determinism and is also an illustration of the dynamics of two other Social Learning theory concepts: Modeling and Vicarious Reinforcement which will be covered in the next section.
One of the most important laws of Social-Cognitive Learning Theory was developed by Rotter in the 1950’s and validated in a number of empirical research studies over a period of two decades from the 1960’s through the 1980’s. It is known as the law of Behavior Potential and is stated in the formula below:
BP = ƒ (E,VR)
It is partially based on Thorndyke’s Law of Effect. Basically, this heuristic formula means that the potential (or probability) of any new behavior is a function (or is dependent upon) two factors, the Expectation the person has of receiving a reward or reinforcement for the behavior and the Value of the Reinforcement expected. Both of these factors must be high for the probability or potential of the behavior to be high. For example if the Expectancy value is .1 and the Reinforcement value is .8, the probability of the behavior being performed will be around 0.08 or less than 10%. If the Expectancy value is .7 and the Reinforcement value is.2, the probability of the behavior being performed will be approximately .14 or less than 20%.
However if both are high (for example, E is .7 and VR is .9), the probability of the behavior being performed is over 60%. This law is relevant to the understanding of social media engagement behavior. Skylab has structured its VRS Gamification system to maximize the positive effect of this law of behavior by insuring that users feel very confident that the behavioral goals and listed actions will be reinforced and thus Expectancy is very high. In addition, Skylab has created a system of meaningful rewards that users value (not just badges but crowd-sourced positive reinforcement for the achievement of goals- e.g., the Recognition Wall, the Leaderboard and Chat options that allow other users to reinforce achievement throughout the multiple experiences of the entire community.
B. Modeling or social learning
Social learning theory draws heavily on the concept of modeling as described above. Bandura outlined three types of modeling stimuli:
1.Live models, where a person is demonstrating the desired behavior or when a Skylab user can see how a particular engagement behavior is reinforced by the community and then becomes motivated to perform similar actions to receive a reward.
2.Verbal instruction, in which an individual describes the desired behavior in detail and instructs the participant in how to engage in the behavior.
3.Symbolic, in which modeling occurs by means of the media, including movies, television, Internet, literature, and radio. Stimuli can be either real or fictional characters.
Exactly what information is gleaned from observation is influenced by the type of model, as well as a series of cognitive and behavioral processes, including:
Attention – in order to learn, observers must attend to the modeled behavior. Experimental studies have found that awareness of what is being learned and the mechanisms of reinforcement greatly boosts learning outcomes. Attention is impacted by characteristics of the observer (e.g., perceptual abilities, cognitive abilities, arousal, past performance) and characteristics of the behavior or event (e.g., relevance, novelty, affective valence, and functional value). In this way, social factors contribute to attention – the prestige of different models affects the relevance and functional value of observation and therefore modulates attention.
Gamification and its relationship to Social –Cognitive Learning
Gamification is the application of game theory concepts and techniques to non-game activities. Game theory is a branch of mathematics (and Social Psychology) that seeks to understand why an individual makes a particular decision and how the decisions made by one individual affect others.
Generally speaking, the overreaching goal of gamification is to engage the participant with an activity he finds fun in order to influence his behavior.
In marketing, a gamification initiative might address the cognitive and emotional aspects of game theory as well as the social ones by including a system of rules for participants to explore through active experimentation and discovery, awarding points or badges for levels of participation, displaying leaderboard scores to encourage competition and offering prizes so that participants have a chance to win something of value.
Every company on earth wants to do a better job at engaging customers and driving employee productivity. It’s serious business for sure but the key to unlocking this may lie in something called gamification.
According to Gartner, gamification is the “broad trend of applying game mechanics to non-game environments such as innovation, marketing,..”. Simply put, applying gaming design principles and techniques can help promote audience engagement for the purposes of solving business problems.
And as it turns out, gamification is on the rise and figures to be a sizable business opportunity. M2 Research projects that the gamification market will grow to over $4Billion in the US by 2018, more than 50% of organizations will gamify the innovation process.
What’s involved in gamification? Gamification generally involve the following elements which are leveraged to solve business-related problems –
Challenge or Quest – every game has a purpose
Attainment – levels of achievement
Status – badges, ribbons, flairs signify status
Competition – leaderboards, scoreboards
Gamification can also be applied in changing employee behavior. Cashiers at Target stores are scored based on the speed with which they successfully check out a customer, aligning both productivity behavior with rewards. Google applies a similar approach to expense control that offers rewards similar to what might be encountered in a video-game. Employees are able to amass points based on travel expense savings which they can then apply toward charitable giving or upgrades on another trip.
This may in fact, represent the next evolution of gamification for the enterprise. Instead of simply applying gamification techniques to a niche problem, companies will actually begin to leverage gamification as part of a behavior management platform where loyalty and engagement can be measured, managed and optimized. Skylab has done exactly this with its gamification platform across all its user groups or tribes.
The Skylab Gamification System
The graphic shown above (originally created by Yu-Kai Chou, 2003) illustrates the components of Skylab gamification platform. The 8 different sectors or nodes have all been developed to conform to the core principles and research findings from the field of Social-Cognitive Learning.
In the context of Social-Cognitive or self-reinforcement, the individual decides which behaviors and rewards are meaningful and which are not. Or another way to describe this is that the individual decides which gamified behaviors have value and are worth pursuing in order to gain a reward (i.e. a badge or other recognition). This is a key tenet of Self-Reinforcement Learning.
In order to optimize value for its users, Skylab provides a wide variety of reinforcements for meaningful behaviors. This is one of the reasons that its engagement levels are significantly above the average for similar sites. The elements of Meaning in gamification theory that are built into Skylab’s gamification platform include the following:
One of the key motivators of self-reinforced learning and engagement is the sense of accomplishment one feels for completing a meaningful task and earning social recognition. This sense of accomplishment is an important part of Self-efficacy ( a key component of Cognitive Social Learning which will be covered in more detail in Section IV.) Feelings of accomplishment and mastery are significant motivational drivers of engagement behavior. The elements of Accomplishment that are built in to Skylab’s gamification platform include:
Badges (Achievement symbols)
Fixed Action Rewards
Step by Step Tutorial
Another essential feature of Self-Efficacy is the feeling of ownership, which is an amplification of the feeling of pride one has for accomplishments and also a feeling of self-worth and authenticity because of understanding one’s unique behaviors, thoughts and feelings. To own your thoughts, feelings and behaviors is also an important feature of an inner locus of control. (Locus of Control is an allied concept of self-reinforcement and will be further discussed in Section V).
It also is one of the building blocks for developing a sense of social responsibility. Ownership also helps one gain a better understanding of one’s worth and improve one’s self-concept and social standing. Developing a valid estimation of self-worth is also a factor in Self-Efficacy-based self-reinforcement.
The elements of Ownership that are built in to Skylab’s gamification system include:
Build from Scatch
In Skylab’s gamification platform, scarcity leads to higher levels of engagement behavior due to the principle of Fixed interval reinforcement. The scarcity function engenders patience, perseverance and a “never give up” attitude which are all important aspects of Self Efficacy and Inner Locus of Control. The following elements are included in Skylab’s gamification system:
Throttles and Moats
Avoidance learning is an important part of any operant-based learning program. Avoidance of a noxious stimulus or a negative outcome is positively-reinforcing and also results in rapid learning of positive and constructive behaviors which is another factor which Skylab’s has created to foster social responsibility. In the Skylab gamification platform, the following elements are included:
Unpredictable reinforcement leads to very strong behavior that is difficult to extinguish and is a real-life example of the Variable Interval reinforcement schedule. In addition, achieving rewards in the context of an unpredictable learning environment also may cause a visceral response and strong physiological reactions which are part of the Self-Efficacy self-reinforcement learning model. The thrill of the chase can be highly reinforcing and lead to increased levels of engagement. In the Skylab model, the following elements of Unpredictability are included:
Social-Cognitive learning leads to increased social influence which is probably one of the most important motivational factors that cause increased engagement. Skylab provides many options for its users to gain social influence and recognition and also allowing leaders to influence the engagement of other users with lesser accomplishments to increase their chances for earning reinforcement because of increased engagement behavior. The following Social Influence gaming elements are included in the Skylab model:
As an individual develops increased Self-Efficacy, feelings of Empowerment are a natural consequence. An empowered person is a highly engaged person and Skylab has created a platform to optimize the positively reinforcing benefits of Empowerment.
The gaming elements that are built in to Skylab gamification model include:
Skylab’s VRS gamification platform has included some of the key elements of the Self-Efficacy model including the following examples:
The Behavior Profile allows a user to see the consistency of their habits through the summaries provided which can be a powerful motivational tool because a user can see how well his/her behavior is tracking to the goals that were set
The Recognition Wall allows each user to be supported in real time by other users and also become aware of whether or not the actions you are taking are aligned with the desired outcomes that have been set. It also provides a user with crowd sourced feedback for taking self-corrective action and developing a better sense of self-efficacy.
The Badges that are awarded for meeting various goals allows a user to reflect upon the journey taken thus far and provides enhanced intrinsic motivation for improved performance
Platform impact dashboard allows a user to see and feel the impact of personal actions and its impact on the overall community so the user can make a reliable judgement of whether or not he/she is making a difference and achieving personal goals.
Skylab is currently planning a series of research studies to confirm that the users are motivated to engage on the platform due to the operation of the various laws and also to measure how much they expect to be rewarded and how much they value the rewards provided for reaching their goals.
This research will also address the average level of self-efficacy attained by users on the different platforms within the Skylab universe of VRS systems.
The Evolution of the Skylab VRS model
This chart illustrates how Skylab USA’s Value Reinforcement System or VRS evolved from Social-Cognitive Learning Theory and its many related theories and concepts. Since its inception, Skylab has gradually incorporated many of the concepts and principles of Social Cognitive Learning Theory as it was originally created separately by the seminal research and thought leadership of Dr. Albert Bandura and Dr. Julian Rotter, two eminent psychologists whose work dates back to the 1950s.