Which statistical test should I choose?

Last updated 2 years ago First published 05 March, 2017

This decision tree will help you to choose the right statistical test for your hypothesis

English Rating: ALL ages Shortest path: 5 nodes Longest path: 11 nodes Possible solutions: 14
Transcription
  • How many data samples do I have? One data sample. Two data sets.
  • What hypothesis do I test? Mean value hypothesis. Variance hypothesis. Median hypothesis. Distribution type hypothesis.
  • Sign test.
  • Has the population normal distribution? Yes.
  • Do I know variance of the population? No. Yes.
  • z-test.
  • t-test.
  • Has the population normal distribution? Yes.
  • F-test.
  • Continuous distribution? No. Yes.
  • Chi-squared test.
  • Do I know values of distribution parameters? No. Yes.
  • Kolmogorov test.
  • Normal or uniform distribution? No. Yes.
  • Lillieforst test.
  • Chi-squared test.
  • What hypothesis do I test? Independence of values. Equal distributions. Variance of the populations. Mean values of the populations.
  • Two sample Kolmogorov test.
  • Contingency table test.
  • Do I have paired observations? No. Yes.
  • Paired two sample t-test.
  • Do the populations have equal variances? Yes. No.
  • Two sample t-test with equal variances.
  • Welch's t-test.
  • Normal distribution? Yes.
  • Paired F-test.

Related Topics

More from this iBosker