The list of software is interesting, but they do almost nothing with it. There is no useful discussion of what these packages do, how to use them, their effectiveness or limitations. Either develop this, put it in supplementary material, or omit it.

While the authors present a table listing nine software packages for SSD fitting, they do not present a detailed review of them. Rather, the focus on just two: the ssdtools R package and the recently released SSD Toolbox (Center for Computational Toxicology and Exposure 2020). It becomes clear that these two packages are highlighted because they offer model averaging, which is some thing they endorse.

13 Responses

  • RESOLUTION

    We have expanded the table on software tools and rewritten the section to allow more comparisons between tools. We now explicitly taste that we focus on ssdtools and the SSD Toolbox because they are the only packages to offer model averaging. Elsewhere in the text we also state that “Several of the current authors are preparing a paper that looks at the performance of the various software tools with various case studies.”

  • Apologies for not pipping in sooner as this was flagged as my item. Thank you Joe for offering to deal with this issue. I think expanding the table is a good idea. I would note that we want to get the message across that this is not just a review paper and that ssdtools is not just “another option” in Table 1 but we are advocating that it is an improvement for many reasons; trying to solve issues with small datasets (model averaging, multi-modal data), using an open-source collaborative platform , and, for us coming from SSDMaster, a more robust statistical method of parameter and CI estimation (MLE and bootstrapping). I think including this type of information in Table 1 and perhaps some additional text would make this clear.

    • Thanks Joe! Some things to add for SSD Master.. Censoring: has an option to truncate- so excluding “tolerant” species in the right tail and only fitting the models to the remaining species. HCx: 1,2,…,98,99. CIs: classic parametric approach. Programming language: Excel,visual basic. Under URL, perhaps its worth noting that it is only available upon request. Analytical method; suggest to spell out terms in a footnote.

  • Perhaps we need to be more explicit about which software tools allow the flexibility to deal with small datasets… see my suggestion in #20.

    • I think this is a distinction worth making but I don’t think it should be the main/sole focus.

  • I raised this issue: I note that in the paper we variously refer to the ssdtools Shiny app and the shinyssdtools app. This is a bit confusing to the reader. Should we not adopt a standard format for this?

    Joe and David both agreed that the latter is appropriate so need to modify the text accordingly

  • As mentioned in my response(s) elsewhere: (i) we are already over the word limits; and (ii) this paper was never intended to be a review of software.

    Kathleen – we could look at writing something for inclusion as Supplementary material as suggested. This could borrow from the review I think you’ve already done.

    • I think we can deal with this by expanding Table 1 to include more information #12 and perhaps moving to supplemental. I’m happy to take this on.

Comments are closed.