1. Week 1
   1. 1.2 Choice Architecture
      1. Humans Are
         1. •Emotional
         2. •Influenced by the context
         3. •Shortsighted and myopic
         4. •Inconsistent
         5. •Imperfect information processors
         6. •Cognitively lazy
      2. Context affects Choice
         1. Compromise Effect
            1. Options tend to get chosen more often when they are the compromise option in a choice set
            2. Simonson, I (1989), Journal of Consumer Research.
         2. Defaults
            1. A no-action default is the outcome when an individual fails to make a decision
            2. Johnson E. and D. Goldstein (2003), Science
      3. Intention – Action Gap
         1. People often fail to accomplish what they want
         2. Short-term vs long-term
      4. Implications of Humans ≠ Econs
         1. Failure rates of new products are high: Perhaps due to a failure to appreciate the underlying consumer psychology
         2. Take-up rates of welfare programs are sometimes low
         3. Policy and Disclosures: The belief that disclosing more information rather than less may backfire
         4. Helping people help themselves – make it EASY for them to get to their outcome, or give them a CRUTCH
            1. Designing contexts [i.e., choice architecture] to help them make better choices [i.e., nudging them]
            2. Providing tools to compensate for human decision-making flaws
      5. Behavior Change Strategies
         1. Strategies non and for-profit
      6. Who Decides?
         1. A higher authority – a government, a policy organization, a think tank of subject matter experts, a self-help group
         2. Social norms, socially acceptable outcomes
         3. The self; the planner within ourselves
   2. 1.3 Rational Choice
      1. 4 C's of Rationality
         1. COMPLETE information
            1. all outcomes/probabilities
            2. all relevant future events
            3. all attributes, even hidden
            4. all relevant context
         2. COGNITION and not emotion driving decision making
            1. emotion is a natural part of human choice
         3. COMPUTATIONAL ability
            1. utility calculation, U=∑wiui(xi)
            2. assumes brain works like a computer
            3. bounded rationality - constrained by computational capacity
         4. CONSISTENCY with the axioms of choice
            1. An axiom is a proposition that is considered to be self-evident.
            2. Axiom: Transitivity
            3. relative comparisons should transfer (if A>B and B>C, then A>C
            4. violated in many cases of human decision making
            5. Axiom: Substitution
            6. If you are indifferent between two options, you should be indifferent between a gamble, which has different properties but the same outcomes.
            7. Cancellation: two options with identical features could be cancelled out if they are opposing
   3. 1.4 Decision Points
      1. Package into smaller units to prompt decisions
      2. more decision points = less consumption
      3. decision points as a nudge
         1. making people save money in small visual envelopes helps people save more
         2. adding pictures of children to the envelope makes the decision point more powerful
         3. makes you stop and think
         4. how to create
            1. transaction costs
            2. reminders
            3. interruption
            4. verification
2. Week 2
   1. 2.1 Mental Accounting
      1. violates fungibility - not all forms of value are equivalent and can be set aside for specific purposes
      2. Mental Accounting Elements
         1. Metering
            1. keeping track
         2. Categorizing
            1. meaningful categories
            2. sources of income
            3. spending categories
            4. time - weekly, monthly, yearly, etc.
            5. individual transactions - set aside for something specific
            6. losing value from a categorized bucket (like losing a ticket you already paid for) feels different than spending new, uncategorized money
      3. Mental Accounting Effects
         1. overconsuming
            1. budgeting too much OR leaving uncategorized
         2. underconsuming
            1. underbudgeting leads to making due
         3. sunk costs
            1. driven to consume what has already been paid for even if they would make a different choice otherwise
            2. weakening sunk cost effect
            3. decoupling
            4. cost per group, not individual physical items
            5. depreciation
            6. time reduces the pain of a past payment
            7. inability to meter accurately
            8. foreign money hurts metering ability
         4. pennies a day pricing
            1. express lump amount as $ per day
            2. compare to other daily expenses
            3. only works if the price is low
      4. Lessons
         1. •When you want people to consume, highlight the cost they have already incurred.
         2. •Conversely, when you would like them to forego consumption, mask the cost. This could be done through transaction formats, timing, or currency.
         3. •Designating money towards a particular purpose makes it more likely to be spent on that purpose
         4. •A match between the source of funds (e.g., cash won in a lottery) and use (e.g., a vacation) increases the likelihood or spending.
   2. 2.2 Choice Overload
      1. observations
         1. Too many choices = no/default decision
         2. potential for regret goes up
         3. leads people to focus on a single attribute instead of making a holistic choice considering all important attributes
      2. Effects
         1. deferral and non-participation
         2. conservative choices (attempt to avoid regret)
         3. status quo, defaults, suggestions have high influence
         4. reliance on peer choices
         5. brand defection - select
      3. combat choice overload
         1. reduce choices
         2. help people understand preferences (bots/agents)
         3. organize options (decision trees, sequential drill-down choices)
         4. attribute-based decision making
         5. outsource to experts (e.g. doctors, financial planners, etc.)
   3. 2.3 Self Control
      1. hyperbolic discounting
         1. illustration
         2. delaying a reward is much more painful than speeding it up is pleasurable
      2. planner vs doer
         1. how can the planner trick the doer into doing the right thing?
         2. Willpower vs Desire
      3. locking mechanisms
         1. social contracts
            1. publicly proclaim your goal
            2. accountability
         2. peer pressure
            1. requires a peer to circumvent the lock
         3. make the default desireable
            1. set aside money in a savings account before ever seeing it
      4. reduce desire
         1. avoidance
         2. distraction or substitution
      5. increase willpower
         1. pre-commit with locking mechanism
         2. impose decision points
         3. cash-based budgeting/spending
   4. 2.4 Spending Uncash
      1. consumption changes as a function of how you pay
      2. metering becomes more difficult the further away from cash
      3. transparency spectrum
3. Week 3
   1. 3.1 Experiment Basics
      1. Why do experiments?
         1. To document behavioural phenomena
         2. To develop a theory to explain them
         3. To study effect sizes
         4. To reconcile and test across theories with conflicting predictions
         5. To test for the efficacy of behaviour-change interventions
      2. Elements
         1. Cause
         2. Effect
         3. Mediator
            1. happens between the cause and effect
            2. example: pennies a day pricing (cause) makes people feel it's more affordable (mediator) resulting in higher spending (effect)
         4. Moderator
            1. modifies the relationship between the cause and effect
            2. example: high prices in pennies a day research inverse the effect
         5. background variables
            1. unmanipulated, natural
            2. results must be interpreted in the context of the background
      3. Techniques
         1. manipulation
            1. anything an experimenter changes across experimental groups
         2. randomization
            1. minimize selection bias
            2. blinding - double blinding is best to eliminate bias
            3. bias can be conscious or unconscious
         3. triangulation
            1. multiple methods to check results
            2. by
            3. dependent measures
            4. methods used
            5. multiple sets of background variables
            6. data analysis techniques
      4. Conditions
         1. control condition
            1. absense of the cause
         2. treatment condition
            1. contains intervention
            2. can have 1 or many levels or permutations
      5. Types of Designs
         1. before and after design
            1. 'natural experiment'
            2. without intervention before, add intervention midway
            3. use separate control group with no intervention for total duration to control for confounding externalities
            4. illustrated
         2. two condition experiment
            1. subjects randomly sorted into control or intervention condition
         3. fully crossed design
            1. total of 4 conditions
            2. illustrated
         4. within or between-participant design
            1. between - different groups get different interventions
            2. within - same group gets different interventions over time
            3. illustrated
   2. 3.2 Analysis
      1. ANOVA
         1. null: the means of all groups are equal
         2. alternative: not all the means are equal
            1. doesn't say how or which ones differ
            2. can follow up with 'multiple comparisons' (AKA contrast tests)
         3. between and within groups variation
         4. F-statistic
      2. Regression
         1. useful when
            1. large number of variables we believe might cause the effect
            2. these variables don't come in neat categories/conditions; might be continuous
         2. draws the line of best fit (least error)
         3. a = intercept
         4. b = coefficient
      3. p-value - confidence measure
      4. null hypothesis
         1. the mean result of control and intervention groups is the same
      5. t-score
         1. difference in means / variability
      6. t-test of differences in means
         1. liklihood the null is false (null hypothesis is rejected)
      7. ANOVA Effects
         1. Main effect
            1. primary effect
         2. Interaction effect
            1. modified effect
      8. Regression Results
         1. look only at statistically significant coefficients
         2. size of the coefficient is indicative of the effect size
         3. interaction effects are captured by including the product of the respective X's
   3. 3.3 Lab Experiments
      1. p value of < 5% means statistically significant
      2. Subtopic 2
      3. Subtopic 3