Discovery -n- Science
| |||
What is discovery?
Discovery is the process of observing something new - either by intention or by accident (serendipity). It can also be the process of observing a new attribute on something that is not new (e.g. the microscope enabled many new discoveries from observations of common everyday objects.) Discovery may form the initial spark for a great scientific discovery, invention or innovation. Alternatively, if there is no follow-up ACTION then it may just remain a discovery - until the next person discovers it.
|
You can even apply the principle of discovery to your very own place of work. By observing things from different perspectives, and looking at the attributes of all those everyday objects and processes, you may well discover things that you were not aware of.
With a bit of practice, you should be able to quickly identify ineffective and inefficient processes that take place almost everyday. If you ask why something is done the way you have observed it, you may well get a reply along the lines of "I don't know" or "because that is the way we have always done it" [which also means they don't know]. Having made those observations you now have an opportunity to make improvements. Equally, you may observe elements of good practice that you were not aware of. You have the opportunity to benefit from those too. |
Science made simple
Science is probably the most robust, and most respected, methodology we have. It is the king of all best practice, especially with regard to discovery.
Science is applied to a vast range of topics from the pure sciences of physics and chemistry to many applied sciences. However, the label is also pinned on some semi-scientific and non-scientific things too. So what is real science (and by deduction what is not)?
Is mathematics a science? - I do not know for sure but, typically, it plays a fundamental role in all real science.
Typically, science consists of the following methodology:
- Observation (perhaps discovery) and measurement
- Hypothesis of the relationship between observable parameters
- Development of a model to represent the hypothesis (or theory)
- Calibration of the model with a set of observed measurements
- Predictions from the model, for a new scenario
- Validation of the model (predictions) against a new set of measurements from the new scenario
If your methodology follows the above steps (and your model is successfully validated) then you can claim that you use a scientific methodology. The model may be a simple mathematical equation (e.g. y = 2x) or it may be a complex computer program (e.g. weather forecasting) - it depends on the complexity of the thing you are modelling. Of course, if the model is not suitable the process can be repeated.
(Models are sometimes associated with what are called laws, particularly in physics (e.g. Newton's Laws of Motion). Some models may not be numerical, for example, they may be based on logic instead. An artist uses a model that predicts when red and yellow paint are mixed the result will be orange.)
So a useful outcome from science is a model of understanding of how a given process works. Such a model can be used to predict and forecast real-world scenarios.
| They say information is power, so what about a model that can predict future scenarios and/or the reaction to a given change? You can use this innovative approach to develop models that are relevant to your organisation. In a business context, for example, this could give you the edge over your competitors, in various ways . . . |
Some sciences are not quite that rigorous though. Sometimes that is because the system being observed is too complex to model in its entirety (e.g. human biology) - but in such cases niche specialisms may help (e.g. molecular biology or bio-chemistry).
Scientifically proven
Now here is a phrase you may have seen many times before: "scientifically proven" - especially in a marketing context! How many of them really went through the above scientific methodology I wonder?
That last step of model validation is the crunch test. The predictions from a model must match the measurements from a new set of observations (and all future observations) if it is to remain valid. But, of course, nothing in life (nothing!?) is black and white, or exact. Predictions from a model have a degree of uncertainty, and measurements have a degree of error too. (If you take a metre rule and measure something to be 54 centimetres is it really 54 or 54.1 or 53.954321?) Similarly, if your measurements involve surveying the views people have, those statistical measurements have an uncertainty too.
For predictions to be valid their range of uncertainty should encompass the actual measured value (including its range of error). For example:
| Prediction | Measurement | |
| Valid (correct) model prediction | 0.90 to 1.10 | 0.98 to 1.02 |
| Invalid (incorrect) model prediction | 0.90 to 1.10 | 0.88 to 0.89 |
In some of the pure sciences (e.g. physics), and engineering disciplines, it is often the case that measurements will also specify the amount of uncertainty, error, or precision associated with it. How many of those marketing slogans seem credible now?
Note that even something that has been scientifically proven, has a level of uncertainty associated with it.
Expect the unexpected
Even very good scientific laws can be invalid when we change the context.
For example, when the context focuses on very small dimensions (e.g. at the atomic level and smaller) the conventional laws of physics break down and the principles of quantum mechanics take over -- which includes behaviour that is totally alien to our common sense. The attributes of an object (e.g. momentum and position) really can be undefined until they are observed, and such observations have a fundamental amount of uncertainty associated with them (no matter how sophisticated your measuring device). Objects can pass through barriers in ways that would seem impossible in terms of common sense. (Objects pass straight through the Earth everyday, just as if it does not exist.) Objects (e.g. electrons and protons) also have the properties of both a particle and a wave!
When the context is changed to one of very high speeds (near the speed of light) the conventional laws of physics break down again. It has been proven that relative to an observer distances really do become shorter and time slows down! When the context is changed to one with strong gravitational fields (e.g. black holes) again conventional physics breaks down. These last two contexts are covered by the theories of special and general relativity - developed by Albert Einstein.
| So back in the "real world" - what is the point and relevance of this? The point is, it illustrates that even with the most sophisticated and best scientific practice, there are always uncertainties and limitations to a given model or theory. The relevance to innovation is that no matter how confident you are, your innovation always comes with a degree of uncertainty and risk. In other words, it can be useful to expect the unexpected and try to develop a contingency plan. |
| Show Page Links |
|
Page Content Discovery -n- Science What is discovery? Science made simple Scientifically proven Expect the unexpected |
Other pages
|
