1) Explanation must be verifiable in further experiments

1) To Ask a question

2) To do background research

3) To construct a hypothesis

4) To test with an experiment

Explanations that explain the correlation between the individual variable and the dependent variable

Explanations that try to identify the truth, and use the truth to explain the cause and the effect in a scientific investigation

For example, scientists in the past used to the "Geocentric Model of the Universe" to explain the natural phenomenon of why people observe stars and other galactic bodies to revolve around us.

Now we know that the model of the universe where everything revolves around the Earth is not true, but nonetheless, this model of the universe helped people explain the phenomenon in "the physical world we occupy"

Another example: Miasma Theory

Scientists in the past used to used this theory to explain why people got sick. Scientists explained that the reason people fell ill were because of "bad air", that spread illnesses to people.

This was a good explanation, because it explained for why more densely populated areas got sick. Scientists said that it was because there was more bad air created from factories.

However, this explanation is untrue, because we now know that illnesses are living organisms, that we refer to as "germs", instead of arising from "bad air".

For example: Germ Theory of Disease

Scientists in the modern day have largely agreed upon this theory of disease, where people fall ill because of microorganisms that attack our bodies. This explanation for falling ill is both true and "good", as it is scientifically proven with experiments and is also good as it is very very good at making predictions.

1) The explanation has to be Justifiable

(It has to be defendable and have proof backing it up)

2) The explanation has to be Verifiable

(As in the explanation must be able to be backed up by new experiments investigating the same issue)

3) The explanation is Testable and Falsifiable

(to hold true to the scientific method, a "good" scientific explantation must be able to be tested, and hence, potentially be able to be disproven)

4) The explanation must be very very good at making predictions

5) The explanation must be transferable

(The explanation must be able to apply to all applicable areas. For example, an explanation for why volcanoes explode must apply to all volcanoes)

e.g. Light is described as a wave when explaining why there are different colours

e.g. On the other hand, light is described as a bouncing ball when describing how the light reflects off a mirror

These two explanations contradict each other, but they both serve as a "good" explanation that accurately predicts real world phenomenon

2) Explanation must not be able to be falsified in further experiments / investigations

3) Explanation must be testable, in order for people to check whether or not they are true

A "good" explanation only has to be an accurate predictor. It must be able to accurately predict the outcomes of experiments that relate to the explanation

While seeking for "truth" may be considered as the ultimate goal for some scientists, "truth" itself is not necessary to make predictions, and therefore is not necessary in a "good" explanation

Analogies are a good example of accurate explanations that provide predictions for an outcome, and at the same time, are not technically "truthful"

Explanations in Natural Sciences must explain for cause and effect, but this does not mean they must be "true".

Explanations must only account for the "how", in disregard to whether the "why" is true or not.