Search results

  1. E

    Test wanted for comparing algorithms

    Q1: Thank you for the confirmation. In this case, I will go with a mixed anova. Q2: Exactly, normality would most probably be an issue. Would it be appropriate to set suitable time limits for each size? If one sets the time limit low enough for each size, one can prevent all responses having...
  2. E

    Test wanted for comparing algorithms

    Ok, that makes sense. Using a suitable software should not be an issue, I just need to understand the study design properly. Question 1: Nature of IV "size" If I am not mistaken, a two way repeated measures anova requires two within-subjects IVs. I have a question about the within-subjects...
  3. E

    Test wanted for comparing algorithms

    Thank you, this is of great help! Could you explain to me your advice was influenced by my answer to your question regarding the scale property of the size? Would you consider size a scale in this context? I will probably need some time to educate me on the two way repeated measures anova /...
  4. E

    Test wanted for comparing algorithms

    Oh, the data is indeed matched, so a paired t-test / Wilcoxon signed-rank test should be used. Thank you! I am not sure about the term "scale" in this context, so let me give you an example. Suppose, we want to find the shortest cyclic tour visiting n cities...
  5. E

    Test wanted for comparing algorithms

    Study design Suppose I want to compare algorithm runtime performance (measured in seconds) of different algorithms for the same problem type. There are several problem instances for this problem type on which the algorithms are tested and performance is measured. Possible tests As far as I...