I'm looking for books and information like crazy and I can not find what I need. Well the example proposed is about methods that have been used in literature students and these are the data collected:

$$ \begin{array}{rr} \text{Method 1, “X"} & \text{Method 2, “Y"} \\ 48 & 14 \\ 40 & 18 \\ 39 & 20 \\ 50 & 10 \\ 41 & 12 \\ 38 & 102 \\ 53 & 17 \end{array}$$

calculate with a level of significance of $\alpha = 0.05$ and P-Value


  • $H_0 : F_{(t)} = G_{(t)}$

  • $H_0 : F_{(t)} \neq G_{(t)}$

The null hypothesis indicates that there is no difference between the reading groups X and Y.

I would like your help but as I don't understand the development well, I would ask you to do it step by step.

and I would also like to know how I solve an exercise where the samples don't have the same size

  • 1
    $\begingroup$ You need to use the self-study tag. $\endgroup$ – Michael R. Chernick Jun 3 '19 at 0:38
  • 2
    $\begingroup$ 1. What's your response variable there? 2. You ask about significance but your question contains no hypotheses; it's only a meaningful question in the presence of them. What's your null and alternative? $\endgroup$ – Glen_b Jun 3 '19 at 3:56
  • $\begingroup$ @Glen_b, edited $\endgroup$ – royer Jun 3 '19 at 16:23
  • $\begingroup$ I presume F and G represent cdfs? $\endgroup$ – Glen_b Jun 4 '19 at 1:15
  • $\begingroup$ @MichaelHardy thanks. I had to go look up how to do that in MathJax. The usual LaTex method didn't work. $\endgroup$ – Glen_b Jun 4 '19 at 1:22

There are several possible tests for this situation.

A common choice is the two-sided two-sample Kolmogorov-Smirnov test.

Note that this assumes that the distributions are continuous; if your variables are discrete this test tends to be quite conservative (lower than desired type I error, with corresponding reduction in power)

As with the one-sample test, the test is based on the largest vertical distance between the cdfs.

This works with unequal sample sizes. There's a table of critical values at the link but any decent stats package will do it for you. You can

There's a small discrepancy between the table in Wikipedia and the p-values provided by R - the table values are large-sample, and are not exact for small samples. (You can find small sample tables fairly easily both in books and on-line.)

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.