I am writing an article critique as part of an assignment on an article titled

Chemical Profiles of Two Pheromone Glands Are Differentially Regulated by Distinct Mating Factors in Honey Bee Queens (Apis mellifera L.)

I am more of a starter in statistical analysis and hence the question. There is a section in the paper titled "Statistical Analyses" in which the author talks about meeting the assumptions of ANOVA. The authors have tested for normality, homogeneity of variances and independence between data collected is assumed. However the author has not mentioned about the 4th assumption of checking if the mean of the residuals are zero.

1) Is this a slip on the part of the authors or is that tested when you draw mean value plots with standard error as they have done in the paper.

My second question is regarding the linear discriminant analysis done in the study. The authors have done an ANOVA to compare between the chemical components of different pheromones and listed the results in a table. I have not used a linear discriminant analysis before, hence,

2) Could you please state why the authors have done a linear discriminant analysis?

3) Also, I am guessing that we use ANOVA for comparison since the comparison is between categorical variables. Is that a correct assumption for this paper?

Any comments please

  • 2
    $\begingroup$ When one performs ANOVA in the usual way, the formulas guarantee the sums of residuals is zero--no checking is needed. Note, too, that this characteristic has a different logical status than the others of the model: the error terms are declared to have expectations of zero in order to make the model identifiable. This loses no generality and it is not something that needs to be checked. It's not an assumption in the same sense that homoscedasticity might be. $\endgroup$ – whuber Apr 14 '17 at 17:06
  • $\begingroup$ @whuber Could you please have a go at the paper, expand on your comment and give it as an answer. I will accept it as the correct answer and close this question. $\endgroup$ – The Last Word Apr 14 '17 at 17:16
  • 1
    $\begingroup$ Thank you, but I see (2) and (3) as the substantive parts of your question, and I have not answered them (nor will I be able to study the paper in sufficient detail to do so). $\endgroup$ – whuber Apr 14 '17 at 17:38

Your Answer

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

Browse other questions tagged or ask your own question.