Test finder

You can either use this text version of a test finder, taken from (Gilbert, McGregor, and Barnard 2017)

or use this flowchart to see which test is appropriate for your study design and your data:

%%| fig-width: 10
%%| fig-align: "centre"
%%| column: page-right

flowchart TD
  A{Difference or Trend Question?} --> B[Trend]
  B --> B2{Are you testing for degree of associationor are you trying to make predictions?}
  B2 --> B33[Association:Test for correlation coefficient Pearson or  Spearman's Rank]
  B2 --> B44[Predictions: Simple linear regression]
  A --> C[Difference]
  C --> C22{Do you have replicates?}
  C22 --> C22Y[Yes]
  C22 --> C22N[No]
  C22N --> C23{Do you have count data?}
  C23 --> C23Y[Yes]
  C23 --> C23N[N0]
  C23Y --> C24(Chi square test: Goodness of fit or Test of independence)
  C23N --> C25[These data cannot be analysed]
  C22Y --> D{How many factors?}
  D --> F[Two or more]
  F --> F2{Independent samples?}
  F2 --> F2Y[Yes]
  F2 --> F2N[No]
  F2Y --> F22Y(n-way ANOVA)
  F2N --> F22N(n-way repeated measures ANOVA)
  D --> E[One]
  E --> G{How many levels?}
  G --> H[One]
  G --> Ha[Two]
  G --> I[More than two]
  I --> J{Independent samples?}
  J --> K[No]
  J --> L[Yes]
  K --> S(Repeated measures one-way ANOVA or Friedman Test)
  L --> T(One way ANOVA or Kruskal Wallis one-way test)
  H --> R(One sample t-test or Mann Whitney U test)
  Ha --> M{Independent samples?}
  M --> N[No]
  M --> O[Yes]
  N --> P(Paired t-test or Signed rank test)
  O --> Q(Two sample t-test or Mann Whitney U test)

Now go to whichever chapter of this book covers the test you need.