Observation Oriented Modeling (OOM)

Overview of Observation Oriented Modeling

• Observation Oriented Modeling (OOM) is an innovative approach to conceptualizing and analyzing data in the social and biological sciences
• It challenges researchers to develop explanatory models of patterns in data
• Research shifts away from aggregate statistics in favor of integrated models
• Focuses on assessing the accuracy of judgements
• Can be taught to undergraduate and graduate students and is adaptable to other sciences
• Requires something of a Gestalt shift on behalf of the scientist, who has been trained in contemporary statistics

Mother and Child Interaction as an Example

• Consider a scientist studying a mother and child interacting in a lab
• Various observations can be made about the mother and child in that artificial setting
• Physical characteristics, height and weight, can readily be measured
• Behaviors, talking and touching, can be directly observed
• Qualities, parenting style or child temperament, may be inferred from other observable behaviors
• Explanatory model is constructed for observations being made, seeking causal structure of natural system under investigation
• In seeking causal structure in nature, the persons, features, behaviors, and qualities are organized or oriented toward one another

Observation and Judgements

• Scientists aren't passive observers, observations are structured based on the scientist's judgements
• Actions such as placing toys and magazines in a room may alter the study
• Decisions made concerning which features or behaviors to look at
• Assess qualities not able to be observed
• Scientist must be aware of assumptions to ensure model captures structure accurately
• Trait scores are generated by summing responses from multiple-item inventories
• Personality traits are structured as quantities and measured as such, even without scientific evidence
• If continuity claim can't be supported with evidence, models must change

Key Actions for the Scientist

• Scientist must steer away from statistical modeling's modern paradigm
• This paradigm is epitomized by structural equation and path diagrams
• Variables connecting are assumed to be continuos quantities, which is based on aggregate statistics (means, variances, etc)
• Judged by estimating pop parameters using null hypothesis significance testing
• Attention is reoreinted to observations and individuals
• Rather than doing statistics or estimating population, patterns are examined in light of a causal model
• Techniques are assumption-free and accurate
• Overall objective: assess accuracy of model to explain observations

Why Observation Oriented Modeling is Needed

-Psychological research has been broken into: -Statistical analysis -Research design -Measurement -Students get more statistical training, though all three are inadequate

• Insufficient training and decline in quantitative psychologists have a negative effect on the overall research
• Modern researchers lack the understanding of statistics, research design and measurement -Published critiques spotlight neglect of scientific investigation principles, replication, critical discourse among colleagues
• Methods drive metaphysics
• Joseph Rychlak argued that psychologies permit their methods to drive their metaphysics
• David Bakan referred to this as 'methodolatry'- nature should conform to research rather than other way around
• Statistics cause neglect of fundamental principles of scientific investigation
• Statistical methods lead to statistical reasoning
• Measurement has been avoided since the early 1900s

Change in General Perspective

• Learning OOM involves "Gestalt-shift" from phenomenalism to moderate realism
• Phenomenalism includes important aspects of positivism and Kant
• Essences of things cannot be known e.g. "things themselves", limited to appearances represented in consciousness
• Personality can be measured as continuous quanities
• Moderate realism states that things have essences (natures)
• Essences may be confused but are a power of the intellect
• Scientists needs to be able to discern different trees
• Student investigates and moves from what is known into territory of uncertienty
• Critical thought, investigation and research lead to conclusion that it should be a shurb

Causation Overview

• Learn to think more about casualty
• Aristotle’s four causes (material, formal, efficient and final)
• Aristotle sought to explain nature
• Can be atoms in the form of chemical compounds Adenine, Cytosine, Guanine and Thymine
• Necessary for an effect

Explanation of Efficient Cause

• Efficient cause likely familiar to students, S->R or SO->R models
• Cause personal happiness as source of change that preceded a certain effect
• To understand the purpose of enrollment and goals that explain human behaviour, final cause needed
• Entails natural end state in order to explain process
• Aristotle reduces humility to hume’s causation and it’s subjective in nature
• Causal explanation is largely a statistical exercise
• Causes bound to natures of things themselves
• Causation in model exists intentionally, the causes “inhered”
• Causes more than humean associations

Initial Exercise in OOM Data Analysis

• Assumed to be continuos qualities
• Reported through the study authors
• Important to transform delay responses

Summary of Accuracy from Study

• Not meant to estimate a population parameter
• Losing power causes observations go un-gracious
• Identify clear and powerful patterns
• Visual rather than w equations though simple equations apply to continous values
• If observations are from self-identified
• Accuracy of integrated models of OOM

Entities/Qualities

• To select analysis from the schemes
• Qualitatively better information through data

Measurement

• Can measure different level of variables
• Distinguishing variables etc can determine relations and relationships
• No data or information is lost by measurement
• Analyzes the OOM for data processing and data sorting

Reviewing the Data

• The data, relationships and data sets through the method
• This information has been gathered by researchers with different aims