Parametric statistical tests assume that your data are normally distributed (follow a classic bell-shaped curve). An example of a parametric statistical test is the Student's t-test.Non-parametric tests make no such assumption. An example of a non-parametric statistical test is the Sign Test.
Parametric tests assume that your data are normally distributed (i.e. follow a classic bell-shaped "Gaussian" curve). Non-parametric tests make no assumption about the shape of the distribution.
Nonparametric tests are sometimes called distribution free statistics because they do not require that the data fit a normal distribution. Nonparametric tests require less restrictive assumptions about the data than parametric restrictions. We can perform the analysis of categorical and rank data using nonparametric tests.
Parametric statistics is a branch of statistics that assumes data come from a type of probability distribution and makes inferences about the parameters of the distribution. See related link.
In parametric analysis the underlying distributions of the variables are described by parameters. These may be known or it may be possible to estimate them from the observed data. In non-parametric analyses, the parameters are not used - either because they cannot be derived or because the tests do not require them.
Parametric statistical tests assume that your data are normally distributed (follow a classic bell-shaped curve). An example of a parametric statistical test is the Student's t-test.Non-parametric tests make no such assumption. An example of a non-parametric statistical test is the Sign Test.
Parametric tests draw conclusions based on the data that are drawn from populations that have certain distributions. Non-parametric tests draw fewer conclusions about the data set. The majority of elementary statistical methods are parametric because they generally have larger statistical outcomes. However, if the necessary conclusions cannot be drawn about a data set, non-parametric tests are then used.
Parametric tests assume that your data are normally distributed (i.e. follow a classic bell-shaped "Gaussian" curve). Non-parametric tests make no assumption about the shape of the distribution.
Nonparametric tests are sometimes called distribution free statistics because they do not require that the data fit a normal distribution. Nonparametric tests require less restrictive assumptions about the data than parametric restrictions. We can perform the analysis of categorical and rank data using nonparametric tests.
parametric
true
Non-parametric statistical methods.
Parametric are the usual tests you learn about. Non-parametric tests are used when something is very "wrong" with your data--usually that they are very non-normally distributed, or N is very small. There are a variety of ways of approaching non-parametric statistics; often they involve either rank-ordering the data, or "Monte-Carlo" random sampling or exhaustive sampling from the data set. The whole idea with non-parametrics is that since you can't assume that the usual distribution holds (e.g., the X² distribution for the X² test, normal distribution for t-test, etc.), you use the calculated statistic but apply a new test to it based only on the data set itself.
You can use statistical tests appropriate for categorical data, such as chi-square tests or Fisher's exact test for associations between variables. For continuous data, you can use t-tests or non-parametric tests like Mann-Whitney U test or Kruskal-Wallis test. It's important to consider the limitations of quota sampling in interpreting the results.
Parametric statistics is a branch of statistics that assumes data come from a type of probability distribution and makes inferences about the parameters of the distribution. See related link.
statistical tests
In parametric analysis the underlying distributions of the variables are described by parameters. These may be known or it may be possible to estimate them from the observed data. In non-parametric analyses, the parameters are not used - either because they cannot be derived or because the tests do not require them.