Advantages Nd disadvantage of non parametric statistics
Answers
Answered by
3
First, let’s look at what nonparametric means. It means not making assumptions about parameters. Parameters are values from the population (as opposed to statistics which come from samples). So, e.g. an independent samples t-test assumes that the means in the populations are normally distributed. Non-parametric tests can still make assumptions (e.g. that the data are independent) but not about parameters.
This shows the main advantage of non-parametric tests: You don’t have to make assumptions that are often dubious or even surely untrue.
In addition, some parametric tests rely on results that are only true with large samples (and “large” is rarely specified exactly). That’s another advantage of non-parametric tests.
The main disadvantages are 1) Lack of statistical power if the assumptions of a roughly equivalent parametric test are valid, 2) Unfamiliarity and 3) Computing time (many non-parametric methods are computer intensive). How much of a problem each of these are varies. 1) Is not as big as a lot of people think. Many NonP tests have almost as much power as their counterparts. 2) May require that the statistician do more explaining. 3) Depends on how much data you have, what sort of computer you have, and so on.
12.2k Views ·
Nonparametric tests refer to statistical methods often used to analyze ordinal or nominal data with small sample sizes. Unlike parametric models, nonparametric models do not require making any assumptions about the distribution of the population, and so are sometimes referred to as a distribution-free methods.
Also this method is used when the data is quantitative but has an unknown distribution, is non-normal, or has a sample size so small that the central limit theorem can't be applied.
Nonparametric tests have some distinct advantages especially when observations are nominal, ordinal (ranked), subject to outliers or measured imprecisely. In these situations they are difficult to analyze with parametric methods without making major assumptions about their distributions. Nonparametric tests can also be relatively simple to conduct.
Disadvantages of Nonparametric methods include lack of power as compared with more traditional approaches. This is a particular concern if the sample size is small or if the assumptions for the corresponding parametric method (e.g. Normality of the data) holds. Nonparametric methods are geared toward hypothesis testing rather than estimation. It is often possible to obtain nonparametric estimates and associated confidence intervals, but this is not generally straightforward. Tied values can be problematic when these are common, and adjustments to the test statistic may be necessary.
This shows the main advantage of non-parametric tests: You don’t have to make assumptions that are often dubious or even surely untrue.
In addition, some parametric tests rely on results that are only true with large samples (and “large” is rarely specified exactly). That’s another advantage of non-parametric tests.
The main disadvantages are 1) Lack of statistical power if the assumptions of a roughly equivalent parametric test are valid, 2) Unfamiliarity and 3) Computing time (many non-parametric methods are computer intensive). How much of a problem each of these are varies. 1) Is not as big as a lot of people think. Many NonP tests have almost as much power as their counterparts. 2) May require that the statistician do more explaining. 3) Depends on how much data you have, what sort of computer you have, and so on.
12.2k Views ·
Nonparametric tests refer to statistical methods often used to analyze ordinal or nominal data with small sample sizes. Unlike parametric models, nonparametric models do not require making any assumptions about the distribution of the population, and so are sometimes referred to as a distribution-free methods.
Also this method is used when the data is quantitative but has an unknown distribution, is non-normal, or has a sample size so small that the central limit theorem can't be applied.
Nonparametric tests have some distinct advantages especially when observations are nominal, ordinal (ranked), subject to outliers or measured imprecisely. In these situations they are difficult to analyze with parametric methods without making major assumptions about their distributions. Nonparametric tests can also be relatively simple to conduct.
Disadvantages of Nonparametric methods include lack of power as compared with more traditional approaches. This is a particular concern if the sample size is small or if the assumptions for the corresponding parametric method (e.g. Normality of the data) holds. Nonparametric methods are geared toward hypothesis testing rather than estimation. It is often possible to obtain nonparametric estimates and associated confidence intervals, but this is not generally straightforward. Tied values can be problematic when these are common, and adjustments to the test statistic may be necessary.
Similar questions