Innovation Analysis

Abstract

The purpose of this paper is to define innovation and to proffer evidence of its existence through the use of statistical method and mathematical modeling. Innovation is defined by the confidence interval of the normal probability distribution density where μ ± 2σ = 0.95 leaving a room of ±α = 0.05 for randomness. This paper claims that the upper tail of the probability distribution density curve ( α = 0.025) is an area of focus for innovation analysis. The lower tail of the curve: −α = −0.025 is a random area that is not a subject of innovation analysis. The paper employs series of tests to determine and verify innovation. Two situations are presented. On the one hand, innovation is tested by comparing the claimed output change against the industry referenced mean. On the other, innovation may also be proven where there is no external reference. Vouching and tracing analyses are used where there is an industry reference indicator. Absent such an external reference indicator, the Dixon outliers test is used.