{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

NonparametricStats

# NonparametricStats - Questions How do you know if your data...

This preview shows pages 1–4. Sign up to view the full content.

Parametric vs. Non-parametric statistical  tests Parametric Non-parametric 2-sample t-test Mann-Whitney test Paired t-test Wilcoxon Signed-ranks or Sign test One-way ANOVA Kruskal-Wallis test (2 or  3 independent  variables) Repeated-measures  ANOVA Friedman test Pearson’s Correlation Spearman Rank  Correlation Parametric tests assume data are normally distributed (bell curve);  have similar variance among groups are measured on continuous scale with equal  intervals;  and

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Parametric vs. Non-parametric statistical  tests are randomly and independently taken from  source population. Use parametric tests when possible; they are  more powerful than nonparametric tests. Use nonparametric tests when: Data are skewed and cannot be transformed. Sample size is too small. Data are ranked.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Questions: How do you know if your data are skewed? Parametric vs. Non-parametric statistical tests o Make a frequency distribution (histogram) of data. o If skewed: Try transforming data 1) Right-skewed data (common in biology): square root(x), log(x) or ln(x), negative reciprocal (-1/x) • Note: Don’t log-transform data with zeroes; first add a number to all values. 2) Left-skewed data: power functions (x 2 , x 3 , x 4 ,…x n ) 3) Proportion data (values between 0 and 1): arcsine(square root(x)) Parametric vs. Non-parametric statistical tests • How small is too small a sample size? o Refer to power analysis for individual tests • What are ranked data? o Data on ordinal scale o Ratings...
View Full Document

{[ snackBarMessage ]}