Exploring Inferential Statistics and Their Discontents - Essay Example

fers to a linear model applied in statistical analysis; it is used in testing hypotheses, factor in estimates, known quantities and other sources of errors in statistical experiments. Thus, GLM matters since it is used in quantifying the relationship between a number of predictor variables and a dependent variable. Parametric methods or statistics refer to those methods that assume normality of a population or used after approximation by normal distribution immediately after invoking CLT (central limit theorem).

On the other hand, non-parametric methods refer to general statistical techniques that do not make assumptions about normality of a population under study. In addition, these methods do not depend on the population which is being studied. Thus, they can also be referred to as distribution free methods. Compared to parametric methods, non-parametric methods are becoming more popular because statisticians or researchers are not supposed to be constrained to make assumptions on the population under study.

Therefore, to compare the two methods, we consider a case of finding confidence interval on the mean. Both methods are required for this type of problem since parametric methods are deemed to be more efficient than non-parametric methods. However, difference in efficiency may not be a big issue but we need to decide on a method that is more efficient depending on the problem in question (Sheskin, 2000). It is crucial to pay attention to assumptions of statistical tests since a number of incorrect background assumptions will definitely result to inaccurate conclusions about a population being studied.

Thus, such assumptions must be made with a lot of care. For cases of dependent variable scores that are not normally distributed, they can be transformed and use the alternative statistical of non-parametric methods in order to make equal weights on the variables. P=0.05 in statistics means the 5% probability of obtaining appropriate results from a given