Outline the Data Generation Process ********* Lesson ================ .. article-info:: :avatar: UCLA_Suit.jpg :avatar-link: https://www.decisionneurosciencelab.com/elijahgalvan :author: Elijah Galvan :date: September 1, 2023 :read-time: 4 min read :class-container: sd-p-2 sd-outline-muted sd-rounded-1 Goal During this Stage --------------- Before we can start to develop an mathematical model of behavior, we have to establish a conceptual model of how :bdg-danger:`Decisions` are made in your scenario that comports with your hypotheses. How to Achieve this Goal ------------ 1. Identify Motives Per Hypothesis - for each hypothesis, think of the central appetitive or aversive motive that would lead people to make such :bdg-danger:`Decisions` 2. Identify Behavioral Patterns Predicted by Each Hypothesis - think about *how* the :bdg-danger:`Decisions` people make enable them to act on their environment ( :bdg-primary:`Independent Variables` and :bdg-primary:`Constants` ) to avoid or obtain whatever is motivating them according to each hypothesis 3. Explain How Other Behaviors Prevent Preferences From Being Met - now think about *how* other :bdg-danger:`Decisions` that people could make would prevent them from avoiding or obtaining whatever is motivating them according to each hypothesis. 4. Identify Dimenions of Differences Between People - taking a step back now, identify the dimensions which are implied by the :bdg-success:`Meaningful Psychological Differences Between People` you have identified in your hypotheses - what way are these people quantitatively different? .. Note:: The central theoretical question we want our Equation to answer is this: What :bdg-success:`Meaningful Psychological Differences Between People` determine how much :bdg-warning:`Utility` they derive from their :bdg-danger:`Decisions` in any situation of interest, where the situation as defined by the :bdg-primary:`Independent Variables` and :bdg-primary:`Constants`. Following these steps should enable you to answer this question in a conceptual way: you now have a conceptual model of how :bdg-primary:`Independent Variables`, :bdg-primary:`Constants` , and :bdg-danger:`Decisions` relate to the :bdg-success:`Meaningful Psychological Differences Between People` identified by each hypothesis. In the next step you are going to translate your conceptual answer to a mathematical one by creating a :bdg-warning:`Utility` equation. Tutorials ================ Tutorial 1 - van Baar, Chang, & Sanfey, 2019 ------------------- .. dropdown:: Identify Motives Per Hypothesis 1. Greed - a desire to obtain money (or an aversion to losing money) 2. Inequity Aversion - an aversion to creating inequity (or a desire to achieve equity) 3. Guilt Aversion - an aversion to violating others' expectations (or a desire to meet others' expectation) .. dropdown:: Identify Behavioral Patterns Predicted by Each Hypothesis 1. Greed - people keep and take as much money as they can 2. Inequity Aversion - people give back enough money so that them and their partner have an equal amount 3. Guilt Aversion - people give back enough money so that their partners' expectations are met .. dropdown:: Explain How Other Behaviors Prevent Preferences From Being Met 1. Greed - not keeping (or taking) as much money as one possibly can does not maximize one's payoff 2. Inequity Aversion - not giving enough money back so that them and their partner have an equal amount results in inequity 3. Guilt Aversion - not giving their partner what they expect results in the violation of another person's expectations which results in feeling guilty .. dropdown:: Identify Dimensions of Differences Between People 1. Greedy people act in their own material self-interest while both Inequity Averse and Guilt Averse people act in others' self-interest 2. Inequity Averse people act in others' material self-interest based on a desire to maintain equality - to be objectively fair 3. Guilt Averse people are motivated to act in others' material self-interest based on a desire to meet expectations - to be perceived as being fair Dimensions: * :bdg-success-line:`D1` - ranges from prosociality (low) to greed (high) * :bdg-success-line:`D2` - ranges from acting based on perceived fairness (low) to objective fairness (high) In our lab, to keep things simple we always refer to :bdg-success-line:`D1` as Θ (Theta) and we associate low values with prosocial preferences and associate high values with selfish preferences. We also always refer to :bdg-success-line:`D2` as Φ (Phi) and we select which social norm it will refer to on an arbitrary basis. .. figure:: dimension_hmtg.png :figwidth: 100% :align: center Tutorial 2 - Galvan & Sanfey, 2024 ------------------- .. dropdown:: Identify Motives Per Hypothesis 1. Payout-Maximization - a desire to obtain money (or an aversion to losing money) 2. Equality-seeking - a desire for everyone to have the same amount (or an aversion to people having different amounts) 3. Equity-seeking - a desire to preserve the preexisting distribution of resources (or an aversion to changing this distribution) .. dropdown:: Identify Behavioral Patterns Predicted by Each Hypothesis 1. Payout-Maximization - people redistribute as a means by which to get as much money as possible 2. Equality-seeking - people redistribute to ensure everyone has the same amount 3. Equity-seeking - people redistribute to ensure that everyone has what they were initially allocated .. dropdown:: Explain How Other Behaviors Prevent Preferences From Being Met 1. Payout-Maximization - not redistributing in order to receive as much money as one possibly can does not maximize one's payoff 2. Equality-seeking - not redistributing as much as possible prevents everyone from having the same amount 3. Equity-seeking - redistributing takes money away from people who deserve to keep it .. dropdown:: Identify Dimensions of Differences Between People * :bdg-success-line:`D1` - ranges from prosociality (low) to greed (high) * :bdg-success-line:`D2` - ranges from equity (low) to equality (high) So *Payout-Maximizers* are characterized at all high values of :bdg-success-line:`D1` while *Equity-seekers* and *Equality-seekers* are characterized at low values of :bdg-success-line:`D1`. Meanwhile *Equity-seekers* and *Equality-seekers* are differentiated by :bdg-success-line:`D2` Tutorial 3 - Crockett et al., 2014 ------------------- .. dropdown:: Identify Motives Per Hypothesis Before answering our question about how harm-aversion changes depending on :bdg-primary-line:`Shock Recipient`, we have to answer a more basic question; namely, **What motivates decision-making in harm-for-benefit situations?** Thus, we have to assess what could motivate behavior in either :bdg-primary:`Condition`? 1. Harm-Aversion - a desire to avoid causing harm 2. Payout-Maximization - a desire to obtain money .. dropdown:: Identify Behavioral Patterns Predicted by Each Hypothesis 1. Harm-Aversion - people always choose the less harmful option 2. Payout-Maximization - people always choose the highest paying option However, based on our task, we may wish to account for a loss aversion. This is because we are giving someone an default value that they must choose to lose out on in favor of something else. To be exhaustive at this stage, we should indeed account for this as it applies to causing more harm or receiving less payout, depending on what the default is. 3. Loss-Aversion .. dropdown:: Explain How Other Behaviors Prevent Preferences From Being Met 1. Harm-Aversion - choosing the more harmful option results causing more harm 2. Payout-Maximization - choosing the lower paying option results in less money 3. Loss-Aversion - choosing an option with either more shocks or less money is less desirable since it causes one to lose out on the option with either fewer shocks or more money .. dropdown:: Identify Dimensions of Differences Between People * :bdg-success-line:`D1` - ranges from harm-averse (low) to payout-maximizing (high) * :bdg-success-line:`D2` - ranges from valuing losses and gains equally (1) to being very loss-averse (high) Tutorial 4 - Li et al., 2022 ------------------- .. dropdown:: Identify Motives Per Hypothesis 1. Inequality Aversion - an aversion to creating inequality (or a desire to achieve equality) 2. Harm Aversion - a desire to avoid causing taking money from others (harming them) 3. Rank Reversal Aversion - a desire to avoid usurping an existing hierarchy .. dropdown:: Identify Behavioral Patterns Predicted by Each Hypothesis 1. Inequality Aversion - people chose a division of resources where both players have equal amounts (or as close to equal as possible) 2. Harm Aversion - people chose a division of resources where neither player loses money (or both players lose as little as possible) 3. Rank Reversal Aversion - people chose a division of resources that maintains the existing hierarchy (i.e. the richer player initially still has more and the poorer player initially still has less) .. dropdown:: Explain How Other Behaviors Prevent Preferences From Being Met 1. Inequality Aversion - chosing a less harmful option produces more inequality 2. Harm Aversion - chosing a more equal option produces less inequality 3. Rank Reversal Aversion - chosing a more equal option reverses hierarchy (when this is possible) .. dropdown:: Identify Dimensions of Differences Between People * :bdg-success-line:`D1` - level of aversion to inequality (higher is more averse) * :bdg-success-line:`D2` - level of aversion to harm (higher is more averse) * :bdg-success-line:`D3` - level of aversion to reversing rank (higher is more averse)