Jaydeep Chakraborty

Data Scientist | Machine Learning Enthusiast

Hypothesis Test
https://github.com/jaydeepchakraborty/py_stat

    Q) How many Samples (Nominal)?
      1) 1 sample
        1.1) parametric
          a) z-OneSmaple Test (known population μ, sd)
          b) t-OneSample Test (unknown population μ, sd)
          So when we should perform the Z test and when we should perform t-Test?
          z: We know the population variance, our sample size is large n ≥ 30
          t: We do not know the population variance , our sample size is small, n < 30
          --> If the sample size is large enough, the sample variance is better estimate of population variance so Z test and t-Test will conclude with the same results.
        1.2) non-parametric
          a) WilcoxonSigned-Rank - OneSmaple Test Test
      2) 2 samples
        2.1) parametric
          a) z-TwoSample Test (known population μ, sd)
          2.1.1) non paired/ independent
            a) Independent t-TwoSample test (unknown population μ, sd)
              a.1) Student's t-test (both population have same variance)
              a.2) Welch's t-test (both population have diff variance)
          2.1.2) paired
            a) paired t-TwoSample test (unknown population μ, sd)
        2.2) non-parametric
          2.2.1) non paired/ independent
            a) Mann-Whitney U Test
          2.2.2) paired
            a) Wilcoxon Signed-Rank Test
      3) more than 2 samples
        3.1) one factor/ variable
          3.1.1) parametric
            a) One-way ANOVA
          3.1.2) non parametric
            a) Kruskal-Wallis Test
            b) Friedman Test ~ (non parametric) repeated measure one way anova.
            ["repeated measure" means that you have (well) repeated measure on the same "subject". ~ paired.
            example - comparing the grades of (3 different groups) students on
            3 different tests will be one way anova.But if we had for each student
            his score for each of the same test, then we could compare between
            the test using the repeated measures taken for each student.]
        3.2) two factor/ variable
          3.2.1) parametric
            a) Two-way ANOVA
          3.2.2) non parametric
    Q) (Categorical)
      a) ChiSquare Test