Individual and Organizational Learning 7 - Summary PDF

Individual and organizational learning 7 Blank slates or innate capabilities? Raffael’s „School of Athens“ A lot of the content of the behavioristic learning theories is still present in todays organiszation Rationalism versus empiricism Rationalism: Knowledge comes from humans „inside“ Empiricism:...

Individual and organizational learning 7 Blank slates or innate capabilities? Raffael’s „School of Athens“ A lot of the content of the behavioristic learning theories is still present in todays organiszation Rationalism versus empiricism Rationalism: Knowledge comes from humans „inside“ Empiricism: Knowledge comes from empiric studies form outside People learn through both and organizations should do the same History of Learning models - Classical conditioning Operand conditioning Social learning Causal learning (Creativity and innovation) Classical conditioning A very short explanation - - - Classical conditioning refers to a learning procedure in which: o A biologically potent stimulus (e.g., food) is paired with a o previously neutral stimulus (e.g., a bell) Co-occurrence of the biologically potent stimulus (US: unconditioned stimulus) and previously neutral stimulus (CS: conditioned stimulus) leads to a specific behavioral response (e.g., salivation in dogs, UR: unconditioned responses) usually similar to the potent stimulus. After a while, the occurrence of the previously neutral stimulus (CS) elicits the identical response as the biologically potent stimulus (UR). Pavlovian dog studies were published in 1897 Together with operand conditioning, these studies form the basis of behavioristic learning theories. Classical conditioning was an accidental finding - - Pavlov was studying physiological responses in the dog’s digestive system. Key aspect of his research: Development of a methodology redirecting the animal’s digestive fluid outside of the body, where they could be measured and analyzed Side effect: o Whenever the technician came to the dogs preparing the meals, the dogs started to salivate. o Why was that the case? o Pavlov was interested and started studying this phenomenon Finding: Dogs must have “learned” to associate the presence of a neutral stimulus with food, which produced the physiological effect. Pavlov’s apparatus Classical conditioning is a phylogenetic old process Examples of Conditioning Eyeblink conditioning Fear conditioning Fear conditioning in humans: Little Albert Studies They scared him with a striking hammer upon a suspended steel bar and documented his reactions (at which point he starts to cry). Operand conditioning (B.F. Skinner) Learning in which strength of behavior reinforced through punishment and reward Structure of operant conditioning -> Abbildung auswendig lernen, kommt an Prüfung Positive Reinforcement: ex. Add someting nice Negative Reinforcement: - avoidance - remove sth. bad Positive Punishment: Add something bad Negative Punishment: Remove something nice Classical versus operant (instrumental conditioning) - - Classical conditioning involves biologically relevant stimuli that cannot (easily/perfectly) be consciously controlled o Examples: Saliva (Speichel), eye blink, fear response Operant conditioning involves “voluntary” behaviors o Examples: Picking for food, avoiding the hot stove, doing overtime work because it leads to a bonus payment or a promotion Prior to operant conditioning: Thorndike’s law of effect (1898) - - Before BF Skinner, Edward L. Thorndike studies operant conditioning intensively Key studies: o Observations of cats trying to escape from puzzle boxes o Manipulation of ease, with which the door unlooked Pleasure of the after-effect strengthens the action with which it is produced o Success experiences spark learning o Continued failure decreases learning Social learning (Albert Bandura: Historical Background) - In the 1940s, B.F. Skinner delivered a series of lectures on verbal behavior, putting forth a more empirical approach to the subject than existed in psychology at the time Proposition of the use of stimulus-response theories to describe language use and development, and that all verbal behavior was underpinned by operant conditioning. o However: some speech is not derived from direct reinforcement... - At around the same time, Hull, a strong proponent of behaviorist stimulus-response theories, and his team at Yale aimed to come up with a reinterpretation of psychoanalytic theory in terms of stimulus-response. o Book: Social Learning Theory (1941) ▪ Personality consisted of learned habits (see drive theory, Lecture 2) ▪ Humans possess a drive of imitation, which is positively reinforced by social interaction and widespread as a result. Historical development: Origins of social learning theory (Bandura) - - Julian B. Rotter, published his book “Social Learning and Clinical Psychology” (1954) o Departure from the strictly behaviorist learning o holistic interaction between the individual and the environment. o In his theory, the social environment and individual personality created probabilities of behavior, and the reinforcement of these behaviors led to learning. o Subjective nature of the responses and effectiveness of reinforcement types. o Use of vocabulary common to that of behaviorism, but strong focus on internal functioning and traits o Pre-curser to cognitive theories of learning. Noam Chomsky published his criticism of Skinner's book Verbal Behavior in 1959 o pure stimulus-response theories of behavior could not account for the process of language acquisition, an argument that contributed significantly to psychology's cognitive revolution. o “Human beings are somehow specially designed to" understand and acquire llanguage, ascribing a definite but unknown cognitive mechanism to it Bandura’s focus on learning of aggression (co-occurrence with TV spread in United States of America) - Albert Bandura studied learning processes that occurred in interpersonal contexts o These could not be explained by theories of operand conditioning o Bandura argued: "the weaknesses of learning approaches that discount the influence of social variables are nowhere more clearly revealed than in their treatment of the acquisition of novel responses.“ o Bandura began to conduct studies on the rapid acquisition of novel behaviors -> neuartige Verhaltensformen -> Bsp. Kinder sprechen Hochdeutsch mit ihrem Vater und Schweizerdeutsch mit ihrer Mutter. Der Vater sagt: „Mir ist kalt“, die Mutter sagt „ig ha chaut“. Die Kinder sagen „mir isch kalt.“ -> In ihnen geht also ein Prozess ab, denn sie haben „mir isch chalt“ vorher noch nie gehört. o Most famously: the Bobo-doll studies Bandura’s social learning theory: Elements - - 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. Key elements: o Learning is not purely behavioral; rather, it is a cognitive process that takes place in a social context. o Learning can occur by observing a behavior and by observing the consequences of the behavior (vicarious reinforcement). o Learning involves observation, extraction of information from those observations, and making decisions about the performance of the behavior (observational learning). Thus, learning can occur without an observable change in behavior. o Reinforcement plays a role in learning but is not entirely responsible for learning. o The learner is not a passive recipient of information. Cognition, environment, and behavior all mutually influence each other (reciprocal determinism). Criticisms of behavioristic learning theories - - Behavioristic learning theories are powerful and are used up to today o Reward/punishment o Behavioral therapy o Incentives systems in economics (see Lecture: Institutional Economics, next week) They completely neglect the “black box” of the mind. What cannot be measured behaviorally, was not considered “scientific”. Psychological mechanisms (emotions, certain cognitive processes such as gain- loss reference points in evaluation) are entirely ignored “Pragmatic” set of theories: “Don’t care why, it just works...” Science: How can we learn how much from so little? - From some peas we learn about the mechanism of heredity o Mendel experiments From bones we learn the existence of dinosaurs o Paleontological findings From some beeping of a machine we learn the existence of gravitational waves o Wave detectors: Laser Interferometer Gravitational-Wave Observatory (LIGO) How can we (humans) learn so much from so little? o The route to knowledge is: Babies Seeing a giraffe only once is enough for a baby, not for an artificial intelligence - Human learners do not need a lot of data in order to make surprisingly good predictions about the world Adult: “Hey, this is a giraffe”! 12-18 months old: “Giraffe” -> They know it after seeing and hearing it maybe twice o Never, lion, dog, etc. Artificial intelligence needs a lot of learning to identify that. Thats the reason why we often have to anwer things like: “Please indicate all pictures that have a road to proceed to this website...” to prove that the decision maker (we) is human. Modern learning theories: The child as a scientist Babies are little geniuses: We are born “knowing” we are human (and not a hospital bed, for example) -> innate capabilities - By age 6 months we know the meaning of many words - Babies at age 15 months use advanced statistical knowledge - Infants at age 16 months attribute failure correctly An organization can perform dramatically better if they do things like a baby – Sebastian Berger Assumptions of modern learning theories - Not everything in the world is chaos There is some objective truth and the statistical signals that we see, hear, and feel give us evidence about the “real” world When there is truth in the world, the probability to receive a signal that reflects this truth is (slightly) higher than the probability to receive a faulty signal o Example: it is not allowed to cross a red light ▪ People sometimes cross the road when the signal is green and sometimes when it is read ▪ The likelihood that someone crosses at green is higher ▪ Hypothesis: We cross when it is green -> Revision if necessary Theory theory and evidence-based management - “Theory theory” suggests that babies learn causal knowledge by continuously coming up with testable hypotheses and with experimental capabilities and statistical expertise to gain causal insights This is exactly what “evidence-based management” aims to do, derive causal knowledge upon which management can act. Evidence-based “causal” learning - - - Evidence-based management originates from evidence-based medicine o Identification o Disseminating o Applying research o Soundly conducted and clinically relevant What else may guide medicine? o Studies show that only 15% of physicians work “evidence-based” ▪ Often use old knowledge ▪ Use routines and “expert beliefs” ▪ Information comes from people with vested interests • Pharmaceutical representatives etc. Manager’s are often not better!!! Sources of “evidence” for management experts - - - Management “gurus” base their advice on o Shakespeare o Jack Welch o Sun Tsu (The Art of War) o Santa Claus o Attila the Hun o Etc. Management problem: o Organizations are much more diverse than humans ▪ Size, market, culture, year, etc. o How can we gather causal knowledge as managers? o How can we learn anywhere near the efficiency of babies? Using the “best available evidence” is a competitive advantage How do we approach that? What passes for wisdom in management? - - - Seasoned practitioners often neglect to seek out new evidence and rather rely on their own experience more than they trust research Most acknowledge the problems of small samples that characterizes their own observation, but they trust more in that information o Persistence of early anchors (-> see bonus system study) Practitioners often rely on the tools that they know best o Particular reward management practices o Particular organizational approaches o Particular motivational strategies etc. Hype and marketing often affects the choice of particular ways of organizing the firm o Agility o Holacracy etc. The excessive reliance on dogma and beliefs - - - - Beliefs about how human are can crucially affect decision making within an organization o Biased belief: People are selfish and rational o Biased decision: We need to have careful control systems in place Data over dogma: o Experimentation can lead to descriptively more accurate views and acting on data can increase performance Ideology can falsely shape management decisions to a much larger extent than empirical knowledge o It can sometimes be hard to argue against ideological managers Pseudo-empiricism o Selective benchmarking o Naïve focus groups (biased, falsely selected) etc. Confirmation bias Creating a culture of organizational learning based on evidence - - Learning (and experimentation) is a mind-set shift in many organizations Causal insights are usually difficult to gather or expensive o Field experiments o A/B test o Econometric modelling o Qualitative social science studies o Inter-and transdisciplinarity o Inductive versus deductive theory building Spending effort on learning (exploration) takes valuable time away from earning (exploitation) Evidence-based management often pays off, but it does not come without investment