Skip to main content

Archived Comments for: Object orientated automated image analysis: quantitative and qualitative estimation of inflammation in mouse lung

Back to article

  1. On number of sections and statistical power

    Vincenzo Della Mea, University of Udine, Italy

    10 February 2009

    After having attended the Authors' congress presentation, I read the paper with much interest, because it depicts very well a protocol for obtaining reliable results from image analysis on histological sections.

    However I would like to point out that the section entitled "Number of sections has little influence on statistical power of the study" reports a result well known since many years: statistical accuracy of a study depends more on the number of subjects enrolled than on the number of samples within the same subject, according to the Gundersen expression “Do more less well” (1). When dealing with rare condition, taking more samples from few subjects seems a shortcut for obtaining statistically significant results, but is not always the case.

    (1) Gundersen HJG, Osterby R. Optimizing sampling efficiency of stereological studies in biology: or “do more less well!”. J Microsc 1981; 121-65.

    Competing interests