Goodness-of-Fit is a numerical measure of how closely a model matches observed data, allowing students or researchers to understand how closely their model reflects reality. The concept is used primarily in statistical analysis when evaluating data from sample populations. The goal of Goodness-of-Fit is to determine how closely the observed data adhere to the modelled data.

Goodness-of-Fit measures essentially how well a model fits the observed data. Use of the Goodness-of-Fit test requires the definition of a critical value, the value a sample needs to have to be considered consistent with the model. The expression of this value is normally the chi-square test. This is a well-known test that divides the degree of freedom of the sample by the square of the difference between the sample and the modelled values.

The Kolmogorov-Smirnov test is also a type of Goodness-of-Fit test. It is used to determine the significance of discrepancies between a sample and population data, when a set of measurements follows a normal distribution (the bell curve). In order for the observed data to be considered consistent with the population data, the maximum difference between the two data sets must be less than or equal to the Kolmogorov-Smirnov statistic. This test will help to determine if the sample comes from a population with a normal distribution or not.

Goodness-of-Fit is a useful tool for researchers and students who want to validate a model. It can provide useful insights into how closely a sample matches a population and may help to narrow down the most likely cause of any discrepancies between the two. Though the concept of Goodness-of-Fit is related to statistics, it can provide real insight into a situation. Understanding if and how closely a model is in line with reality is beneficial and can provide alternate paths of exploration in the case of unexpected results.