Logic
Scientific Method
How do we know that aspirin works, and homeopathy doesn't? How do we know the Earth is 4.5 billion years old, and not 6,000? Science is not just knowledge - it is a method for obtaining reliable knowledge. Understanding this method protects against pseudoscience and manipulation.
- **Medicine:** Clinical trials follow strict scientific method protocols. Before a drug reaches pharmacies, it passes through several phases of testing. Understanding this helps you evaluate medical news.
- **Nutrition:** Every week a new study says 'coffee is harmful' or 'coffee is beneficial'. How to assess these studies? Sample size, control group, funding source - all matter.
- **Psychology:** The reproducibility crisis showed that many popular effects (power pose, ego depletion) failed to replicate. Understanding this helps not to blindly trust TED talks.
Scientific Method
**The scientific method** is a systematic procedure for obtaining reliable knowledge. It is not a rigid algorithm but a set of principles: observations lead to hypotheses, hypotheses are tested through experiments, experiments are reproduced and criticized. The key property: conclusions must follow from evidence, not from authority or intuition.
**The cycle of scientific knowledge:** Observation → Hypothesis → Prediction → Experiment → Analysis → New hypothesis. This cycle never ends - each answer raises new questions. Science is not a collection of final truths, but a process of ever more accurate approximation.
Why does science require reproducibility of results?
Hypothesis
**A hypothesis** is a testable explanation for an observation. Not every guess is a hypothesis. A proper scientific hypothesis must be: specific (what exactly is being claimed), falsifiable (it can be proven false), and make predictions (if true, what should we observe?)
**Difference between hypothesis and theory:** A hypothesis - an initial assumption requiring testing. A theory - a well-established explanation supported by many experiments and evidence. 'Theory' in science does not mean 'guess' - it means 'most reliable explanation.'
Which of the following is a scientific hypothesis?
Experiment Design
**Experimental design** determines whether the results will be reliable. The main principles: control group (comparison with baseline), random assignment (to remove selection bias), blinding (to remove observer bias), adequate sample size (to obtain statistical power).
**Types of experiments:** 1) Randomized controlled trial (RCT) - gold standard in medicine. 2) Natural experiment - when nature itself varies the conditions. 3) Observational study - when intervention is impossible. Each type has its own limitations.
An experiment compares two groups: one takes a supplement, the other does not. But both groups know who takes what. What problem does this create?
Reproducibility
**Reproducibility** - the ability to repeat a study and obtain the same results. This is the cornerstone of scientific reliability. The 'reproducibility crisis' of the 2010s showed that many published results (especially in psychology) failed to replicate. This is not the end of science - it is science correcting itself.
**Reasons for the reproducibility crisis:** 1) p-hacking - selecting significant results from many tests. 2) Publication bias - only positive results are published. 3) Small samples - insufficient power. 4) HARKing (Hypothesizing After Results are Known). 5) Measurement errors.
Science always gives definitive answers
Science gives probabilistic answers that update with new evidence. 'The data suggest' is not weakness - it is honesty
Absolute certainty belongs to mathematics and logic. Empirical science always deals with incomplete information. 'We are 97% confident' is not doubt - it is precise quantification of confidence.
A landmark study failed to replicate. What does this mean?
Key Ideas
- **The scientific method** is a cycle: observation → hypothesis → prediction → experiment → analysis. Each round refines knowledge.
- **A hypothesis** must be specific, falsifiable, and make testable predictions. 'Vague' and 'untestable' are not scientific.
- **Experimental design** determines reliability: control group, randomization, blinding, adequate sample size.
- **Reproducibility** is the cornerstone of science. The reproducibility crisis is not a collapse of science - it is science correcting itself.
Related Topics
The scientific method is the foundation of empirical knowledge:
- Falsifiability — Popper's criterion: a theory must be testable
- Causation — How to distinguish cause from correlation in experiments
Вопросы для размышления
- How would you test the hypothesis 'social media reduces concentration'? Design an experiment - what is your control group, what do you measure?
- Why is it difficult for scientists to publish negative results ('there is no effect')? How does this distort the overall picture of knowledge?
- Is it possible to apply the scientific method to social phenomena - friendship, love, justice? What are the limitations?