Lifestyle Management
through System Analysis
 

Causal Models

This lecture shows you how to create a causal model of your behavior. 

Objectives

1. Prepare a causal model
2. Conduct imaginary experiments with your causal model
3. Conduct thought experiments on relationship between causes and effects
4. Analyze the effect of multiple causes and constraints on your target behavior

Assigned Reading

1. Reading for lecture on Self Experimentation & Causal Modeling.  More►
2. "Chapter 4:  Causal Models" and "Chapter 5:  Observation versus action" in Steven Sloman  How people think about the world and its alternatives.  Oxford University Press, 2005.  Buy►

Presentations

1. Overview  Slides►   Slides 2003► 
   1. Make a comprehensive list of causes and constraints that might affect your behavior.  Slides►  Slides 2003► 
   2. Make a causal model  Slides►  Slides 2003► 
   3. Conduct thought experiments to verify the assumptions  Slides►  Slides 2003► 
   4. Maintain a causal diary  Slides►  Slides 2003►
   5. Analyze the diary  Slides►   Slides 2003►
   6. Take action & repeat the process  Slides►   Slides 2003►

What you know?

Advanced learners like you, often need different ways of understanding a topic. Reading is just one way of understanding. Another way is through writing about what you have read.  The enclosed assessment is designed to get you to think more about the concepts taught in this session. 

1.  Through which of the following steps can a person find causes of his/her behavior? 
   1. Repeated  observation of causes and effects
   2. Experimentation (observing the effect when the cause is present or absent)
   3. Imaginary experimentation (thinking through the effect when the cause might be absent)
   4. None of the above
2. Using the following information calculate the maximum and minimum impact of scheduling exercise time on exercise.  Assume that on unusually busy days the schedule is ignored. 
   

   Day	Unusually busy days at work	Schedule exercise time in calendar	Exercise
   1	1	1	0
   2	0	0	1
   3	0	0	0
   4	0	1	1
   5	1	1	1
   6	1	1	0
   7	0	0	0
   8	0	0	1
   9	0	0	0
   10	1	1	1
   11	0	0	0
   12	0	0	1
   13	1	0	1
   14	0	1	1
   Table 1:  14-day Diary for Causes of Exercise 1=Yes, 0=No

Assignment Due this Week

1. Every week ask a question or comment on the lecture.  Comment►  Ask►

2. Create and test a causal model.  This assignment is due at end of semester but the outline of your final report and draft of any sections that you have completed should be emailed to the instructor.  The sections where most students will have completed by end of this week are in bold.   The project's final report should have the following sections (see examples in the reading section and within the slides):

   1. Title and contact information

      • Title

      • Name

      • Email

      • Acknowledgement

      • Class-specific student ID

   2. Comprehensive list of causes and constraints

      • List of causes and constraints and dates they were added to the list (in case some were added after data collection started)

      • A discussion why each cause/constraint meets the requirements for causes.  Please discuss each criteria for each cause. 

   3. Causal model

      • The network model of causes leading to the target event.  Create a visual model using any software accessible to you (e.g. Power Point) or draw by hand and scan it into an image and insert into your report.

      • The hypothesized constraints within the model.  List each constraint and which cause is affected.

      • The hypothesized relationship among the causes.  Draw a line between any two causes that are likely to be associated with one another.  Write about the association among the causes.

   4. Thought experiments

      • List of the serial nodes observed with the network and think-it-through experiments (logical statements that support your views) that verified the assumptions for serial causal relationships

      • List of nodes where multiple causes have same effect.  Please show that you have thought through the assumptions behind these types of nodes and that these assumptions are reasonable.

      • List of nodes where a common cause leads to different effects.  Please show that you have thought through these assumptions and they meet the requirement for common-cause node structures

      • List out counterfactual tests of various causes.  For each cause, list examples of situations where the counterfactual statements are reasonable.

   5. Maintain a causal diary

      • Keep data for at least 2 weeks and show that you have kept data not only on your success and failures but also on the presence or absence of various causes.  Please make sure that on each day you indicate both the causes that were present and those that were absent.

   6. Analyze the diary

      • For each cause, calculate the maximum and minimum probability of observing the effect when the cause is present

   7. Take action

      • Describe how you used the insights gained through this analysis to change your lifestyle

   8. Introspection

      • Discuss whether the procedure of making a causal model and testing it is useful.  How could this procedure be made more useful.