# The Value of Imperfect Science: An Examination of Utility
Written on
Understanding the Limited Scope of Science
In contemporary discourse, the standing of science has been undermined by ideological and political agendas. This shift in perception arises from science's inherent nature of falsifiability, the tentative nature of its claims, and the statistical basis for its truths. Consequently, science struggles to match the absolute certainty, clarity, and decisiveness often associated with political ideologies or dogmas, particularly on contentious subjects like climate change, vaccination, intelligence, race, and gender identity.
Advocates of science may too quickly dismiss theories, models, and evidence that do not meet stringent criteria for statistical validity or the accepted definitions of what constitutes scientific inquiry. Science is revered as a beacon of truth in a world filled with misconceptions, but anything falling short of rigorous standards is often deemed worthless or worse.
This raises the question: Is there value in scientific approaches that may not be as concrete as the hard sciences like physics or chemistry? In this discussion, we will define science, examine various models and disciplines that may not fit traditional definitions, and assess their potential utility.
Defining Science
Philosopher Karl Popper succinctly defines science as that which is falsifiable. While the implications of this definition extend far beyond this brief analysis, it serves as a vital guiding principle when evaluating the potential worth of ideas that may not align perfectly with conventional scientific standards.
The Role of Non-Scientific Fields
While science serves as a powerful, enlightened lens through which to view the world—free from cultural or political biases—we must guard against "scientism," which can manifest in two primary ways:
- Scientism as Overreaching Claims: This occurs when broad scientific assertions are made based on limited or weak evidence. For example, deriving widespread conclusions from a single study with a small sample size may lack statistical significance.
- Scientism as Exclusive Inquiry: This perspective holds that only scientific methods are valid for inquiry, dismissing philosophical, moral, or legal considerations.
For our purposes, we will focus primarily on the first definition of scientism, examining the validity of scientific assertions, though the second definition—concerning science's exclusive role in inquiry—is equally intriguing.
To illustrate, consider the relationship between intelligence quotient (IQ) and human value. While IQ is a scientific construct, equating it with moral worth conflates distinct realms of understanding—one rooted in empirical science and the other in philosophy or ethics. This muddling of scientific and ethical issues is prevalent in research and public discourse, often leading to contentious debates.
Clarifying Explanation and Prediction
Within the scientific domain, distinguishing between models used for explanation and those for prediction is crucial:
- Explanatory Models: These are retrospective, statistical, and descriptive, commonly found in softer sciences like psychology, economics, and sociology, as well as in harder sciences.
- Predictive Models: These are forward-looking, probabilistic, and prescriptive, more common in empirical areas like machine learning, which may not adhere to rigorous scientific frameworks.
Some disciplines, such as physics, can serve both explanatory and predictive roles. However, confusion arises when explanatory models are misapplied as predictive tools. For instance, power-law distributions are often observed in various fields, such as:
- The frequency of earthquakes compared to their magnitudes.
- The size distribution of companies.
- The casualty figures in wars.
In these instances, the key takeaway is that while larger, more extreme events occur less frequently, smaller events are more common. It's vital to recognize that power laws, which excel at explanation, should not be indiscriminately used for prediction.
Establishing the Domain of Validity
All scientific claims must define their domain of validity. Here are some illustrative examples:
- The effects of drugs tested on mice cannot be directly applied to humans.
- Newtonian mechanics fails to account for objects moving at relativistic speeds, which require special relativity, while atomic-scale phenomena necessitate quantum mechanics.
- Scientific models developed for small groups may not apply to larger populations.
Hard vs. Soft Sciences
Science is broadly classified into categories:
- Hard Sciences: Fields like physics and chemistry, where rigorous mathematical frameworks allow for robust modeling and predictions.
- Soft Sciences: Disciplines like psychology and economics, which rely on less stringent theories and mathematics, often suited primarily for explanation.
Occupying a middle ground are fields such as genetics and neuroscience, which may possess clearer results than soft sciences but lack the precision of hard sciences.
Some prominent figures, like Nassim Taleb, author of "Fooled by Randomness" and "The Black Swan," argue vehemently against the validity of soft sciences, claiming they can be misleading or dangerous if misapplied in policy-making.
While I appreciate Taleb's perspective, particularly from my background in physics, I aim to delve deeper into psychology as a case study. I will analyze the Myers-Briggs Type Indicator (MBTI) and the Big Five personality traits in light of scientific definitions, their utility, and their predictive versus explanatory capacities.
Exploring Psychological Models
The Myers-Briggs Type Indicator (MBTI)
Developed in 1943 by Katharine Cook Briggs and her daughter Isabel Briggs Myers, the MBTI is based on Carl Jung's theories regarding human cognition, preferences, and lifestyle choices. While Jung's elaborate cognitive functions underpin the MBTI system, we will focus on the four clear dimensions it outlines, leading to sixteen distinct personality types:
- Introversion (I) vs. Extraversion (E): Preferences for solitude versus social interaction.
- Intuition (N) vs. Sensation (S): How individuals perceive the world—through internal models versus sensory experiences.
- Thinking (T) vs. Feeling (F): Decision-making processes—objective logic versus subjective consideration of others.
- Judging (J) vs. Perceiving (P): Approaches to the world—incorporating external structures versus flexibility.
Critiques of the MBTI
The MBTI has faced significant criticism, often labeled as pseudoscience or "psychological mumbo jumbo." Critics argue that replicating scores across different testing sessions is challenging, and results can vary based on time or personal circumstances. Additionally, there is no comprehensive theory unifying the MBTI dimensions with biological or neurological foundations, casting doubt on its scientific validity.
Salvaging Value from MBTI
Assuming the MBTI is entirely pseudoscientific, can we still extract value from it? Perhaps we can consider its dimensions as qualitative descriptors of personality rather than predictive indicators of behavior. For instance, describing someone as introverted or extroverted can convey meaningful information about their social preferences.
The Big Five Personality Test
Many psychologists favor the Big Five personality model over the MBTI, arguing that it offers a more empirical framework. The Big Five, remembered by the mnemonic OCEAN, comprises:
- Openness to Experience: Inventive vs. consistent.
- Conscientiousness: Organized vs. easy-going.
- Extraversion: Outgoing vs. reserved.
- Agreeableness: Compassionate vs. detached.
- Neuroticism: Sensitive vs. secure.
While the Big Five lacks an underlying theory of the mind, it is grounded in empirical research, distinguishing it from the MBTI's non-empirical roots.
Mapping MBTI to the Big Five
Assuming the Big Five is more scientifically valid than the MBTI, we could explore potential correlations between their dimensions:
- MBTI: Introversion vs. Extraversion ↔ Big Five: Extraversion
- MBTI: Intuition vs. Sensation ↔ Big Five: Extraversion and Conscientiousness
- MBTI: Thinking vs. Feeling ↔ Big Five: Neuroticism and Agreeableness
- MBTI: Judging vs. Perceiving ↔ Big Five: Openness to Experience and Conscientiousness
These mappings are not definitive but serve to illustrate how a non-scientific model might correlate with a more empirically grounded one.
Conclusion: The Utility of Non-Scientific Models
In summary, we have examined the essence of science and acknowledged its strengths while recognizing that it is just one of many perspectives for understanding the world. By treating the Myers-Briggs Typology Indicator as a descriptive tool rather than a scientific model, we explored the more widely accepted Big Five personality assessment. Additionally, we attempted to find correlations between these two frameworks.
This analysis suggests that even non-scientific models can provide valuable insights when approached as descriptive rather than predictive tools. This principle can extend to various scientific fields, provided we remain cognizant of their limitations.