The fundamental statistical concepts often work as pre requisites for a thorough understanding of the parametric and non-parametric tests. The fundamental of statistics include the understanding of random variables, probability distributions, parameters, population, sample distributions and the central limit theorem.
The existence of statistics is largely because it is not possible to gather and compile data from the entire population. The solution, so as to ensure the generalisation of research is to collect data from a subset of the population so as to know the best possible truth about the population.
The different quantities such as mean, standard deviations and proportions are all significantly important values and can be termed as parameters in context to the study pertaining to the entire population. For the reason that it is not possible to get data for the population as a whole the estimated quantities from the chosen sample, with the aid of these parameters do help in generalising the analysis.
The parametric procedures assume and rely upon the distribution of the population being normal and also the form of the parameters being normally distributed, such as means and standard deviation). On the contrary, the non-parametric procedures in statistics do not depend upon any assumptions related to the shape of the population or the form of the parameters from which the sample has been extracted.
It does sound appealing and simpler to adopt the non-parametric procedures because of the few or no assumptions they make about the population distribution from which the sample has been extracted. However, they do have two main drawbacks. The main drawback is that much less powerful statistically as compared to their parametric counterparts. The term less powerful means that there are lesser chances of the procedure being able to conclude that association between two variables when they are actually truly associated. A non-parametric test will call for a larger sample to be drawn out in order to have more concrete results.
The second drawback that has been associated with non-parametric tests is that the results of these tests are all the more difficult to interpret as compared to the parametric tests. Even a novice researcher may be able to analyse and interpret a parametric procedure analyse but it certainly calls for an experienced researcher to look into and analyse Non parametric results.
