Thank you very much for this excellent tutorial Please tell me how we can do ANCOVA for the height and weight relationship for male and female to understand any significant difference
I have two questions What can we do ıf levene test significant at .001? (How to handle with for contiuninig analysis with referance ) Second you only tested two assumptions related to Ancova however some texbooks say more assumptions (e.g Field (2009, p. 418) with my best wşshes thanks for your help as you say good luck yor research project
+Ertuğrul Şahin ANCOVA (and it' post-hocs) is typically robust enough to handle nonequality of variance but should be noted in your results.I discuss checking for the presence of 5 different assumptions. According to Field (4th edition) I have covered the primary assumptions.
Many thanks for this clear tutorial, but I have a question that, what should i do next if I find interaction between covariate and factor is significant. should i stop this ANCOVA ?
I am not sure about the covariate. If gender is independent which is a fixed factor, then can seasons can be taken as fixed factor ? What I want to test is that I have collected male and female fish in each season and each male and female fish has got respective length and weight in each seasons. for example:-. dry season wet season male female male female lg (cm) wt(g) lg. wt. lg wt. lg. wt 2cm 3 g 2 3.1 3.2 4 2 2.5 .....etc. Now, I want to test if there is any significant difference between male and female fish between the seasons and male and female fish in each season.
Simon Jango Yes, season and gender could be fixed factors. If that is the case, you don't need to do an ANCOVA, you would perform a factorial (2-way) ANOVA.
Sir, I have a doubt. I am taking seasons and gender as fixed factor, weight as dependent variable. with this I can perform two way ANOVA and get the output. The doubt is length required for the output ( in two ANOVA length is not required or What should I do with length and if I take length as covariate then it is ANCOVA).
