Module 1 Article: From Natural Philosophy to Modern Science

The big idea

Research is a human practice with a long history. It began as "natural philosophy" (reasoning from first principles), evolved through Bacon's experiments and the Royal Society, was mathematized by Newton and later by statistics (Fisher), and today spans Big Science, Big Data, and open science. Understanding this history explains why we do experiments, why we randomise, and why we peer-review.

A timeline of research

Period	Key ideas and figures
Ancient Greece	Aristotle's logic; observation without experiment
Medieval	Preservation of texts; early experimental practice
Renaissance (15th–16th c.)	Experimentation (Galileo); printing spreads knowledge
Scientific Revolution (17th c.)	Bacon's Novum Organum (empiricism); Newton's Principia (laws + maths)
18th–19th c.	Specialisation; the scientific journal; peer review
Early 20th c.	Statistics: Fisher (ANOVA, randomisation), Neyman–Pearson (hypotheses)
Mid-20th c.	Big Science (Manhattan Project, space); research institutions multiply
Late 20th c.	Interdisciplinarity; ethics review (Belmont Report); computing aids
21st c.	Data-intensive science; open access/data; reproducibility crisis; AI-assisted discovery

The four pillars of good research

Good research is:

1. Systematic — a planned, ordered procedure.
2. Controlled — variables are managed (control group, randomisation).
3. Empirical — evidence from observation/experiment.
4. Logical — conclusions follow from the evidence.

Plus two essentials: replicable (others can repeat it) and ethical (honest, with integrity).

Types of research

Type	Goal	Example
Basic	Understand "how/when/why"	Discovering a new algorithm
Applied	Solve a practical problem	Using algorithms to optimise bus routes
Quantitative	Measure and test with numbers	A/B test of two UI versions
Qualitative	Explore meaning and experience	Why do students drop a course?
Mixed	Combine both	Survey (numbers) + interviews (meaning)

Philosophy of research — three dominant views

This is the deepest fork in the road; it determines whether you will:

• test a single objective truth (positivism),
• assume truth is approximate and your methods are fallible (post-positivism),
• or explore reality as co-constructed by people (constructivism / interpretivism).

For our running study (exercise vs stress)

• Positivist: stress is an objective quantity; measure it precisely with the PSS-10; randomise, measure, infer.
• Post-positivist: the PSS-10 is an imperfect proxy for "true stress"; results are provisional and open to replication.
• Constructivist: "stress" means something different to each participant; we explore their lived experience through interviews.

Interdisciplinary and transdisciplinary trends

• Multidisciplinary = different fields work alongside each other.
• Interdisciplinary = fields combine methods to tackle a shared problem.
• Transdisciplinary = fields merge into genuinely new frameworks.

Modern challenges (climate, pandemics, AI) almost all demand transdisciplinarity: epidemiology + computer science (COVID dashboards), psychology + data science (behavioural insights), biology + ML (protein folding).

Case study: the reproducibility crisis as a research-evolution event

Many "significant" findings from the 2010s failed to replicate. The response — pre-registration, open data, larger samples — is itself research evolving. It shows that the practice of research improves when the community self-corrects.

Exam angle

For "history and evolution of research":

1. Name 3–4 milestones (Bacon's empiricism → Newton's maths → Fisher's statistics → 21st-century open science).
2. Link each to a feature of modern research (experiment, maths, stats, openness).
3. Define the four characteristics of good research.
4. For philosophy, contrast positivism (one truth, observable) with constructivism (socially constructed reality).

See also

• Module 10 (ethics) and Module 9 (writing) for the modern reforms above.
• Glossary.md (research, validity, paradigm, ethics committee).