In the field of scientific study, there is a distinct and essential process that every project must go through before the results can see the light of day. First, you’ll start with a hypothesis; an idea based on experience, analysis and intuition. Next, you’ll need data – as much high-quality, varied information as necessary to generate the results you’re looking for. Once you’ve got your data you can begin compiling results looking for findings, but when do those findings stop being concepts and become verifiable facts?
In science, P-values are used to determine the likelihood that a hypothesis is correct. Once all of the data for a study has been gathered, scientists will be able to see that the results point to one or more conclusions. By determining the likelihood that the data that leads to said conclusions has been thrown up by chance, or by a statistical anomaly, they are able to assign their findings a P-value. The likelihood that there is no correlation between the data and the conclusion is known as ‘the null hypothesis’ while the likelihood that there is a correlation is known as ‘the research hypothesis’ – the difference between the two is how scientists look to determine the P-value of a given discovery.
In the world of physics, few discoveries are ever announced to the press without having first achieved a P-value of 5 Sigma, or a 1 in 3.5million chance that the suppositions made based on the data in question is incorrect. Scientists have been bitten before by releasing results based on 3 Sigma P-values which have subsequently been found to have been statistical anomalies, so 5 Sigma is held by most as being the only sure-fire way of guaranteeing the veracity of their findings. The recent discovery of the Higgs Boson particle at CERN, for example, was released because the findings subscribed to a 5 Sigma Standard deviation. In this way, scientists are able to turn that which was previously mere ‘supposition’ into scientifically accepted ‘fact’.
While your study, research project or internal audit may not be of sufficient scientific moment to warrant a P-value, we’re 5 Sigma-sure that the results will mean a great deal to your business. Take a look at our range of data capture services to ensure that the data you capture is top quality and far-reaching, reducing the chances of you making an erroneous assumption from your findings.