Reproducibility
Repeat the experiment. If a hypothesis is true, the results of an experimental test must be reproducible. One negative outcome falsifies the assertion.
Double Blind Experiments
We are biased. Both the experimenter and subjects must be unaware of the hypothesis that is being tested to avoid biases.
Keep all the Data
Data trimming. Keep all of the data. Data that falsifies a hypothesis must not be ignored. Even well trained scientists have been known to subconsciously trim data that did not fit their favorite theory.
Appeal to Authority or Hearsay
Eyewitness accounts are unreliable. It is common for many people who witness the same event to give contradictory reports. Experiments should be designed that do not rely on second-hand accounts.
Authority. Evaluate a claim based on the evidence, not on the pronouncement of an authority.
Ad Hominem Arguments's. Don't judge an argument false based on the reputation of the individual presenting it. For example, don't discount a conservative when she presents scientific evidence, nor an atheist when making a point about morality. Evaluate the argument based solely on the quality of the evidence.
In this day and age, it is becoming more complex for a layperson to evaluate scientific or technological issues. So, we must at times rely on experts. In scientific disciplines, theories that are accepted by a large majority of scientists are usually a safe bet. When science works properly, the consensus will change when new evidence surfaces that contradicts the status quo.
While the special theory of relativity proposed by Einstein and the theory of continental drift proposed by Wagner were both well outside the domain of contemporary theories, scientists accepted them as evidence accumulated. |
Correlation and Causality
Correlation does not imply causality. A correlation between two events is not proof that one event is the cause of the other.
Preponderance of Evidence
Favor the hypothesis with the strongest supporting evidence. Nothing can be proven beyond a shadow of a doubt. It is possible that aerodynamic laws are incorrect, but when a plane crashes, we rightly assume that pilot error, bad weather, or mechanical problems are the most likely causes.
When study after study shows no evidence for the efficacy of prayer; but, studies repeatedly show the success of medical treatment, it is likely that medical treatment is the cause of recovery in patients.
Confirmation Bias
Fooling ourselves. The human mind amplifies the evidence that supports what it wants to believe and conveniently ignores the contradictory evidence. This makes it important to always apply the scientific method when seeking the truth.
It feels good. Functional MRI studies show that we get a high from a confirmation of our biases. If it feels good, don't be swayed by your emotions.
Change your Mind!
When new evidence contradicts your world view, simply change your mind. The electorate and politicians seem to think that changing one's mind is a bad thing. When seeking the truth, a mind change is a required part of the journey. |