Reader Comments

Post a new comment on this article

Referee comments: Referee 1 (Linus Andersson)

Posted by PLOS_ONE_Group on 20 Mar 2008 at 17:45 GMT

Referee 1's review (Linus Andersson):

REVIEW
Tubaldi et al.
The Grasping Side of Odours

This study assessed the relation between the perceived "size" of odors and subsequent reaching and grasping movements. This line of inquiry is interesting and of scientific relevance, as it addresses questions regarding olfactory representations and how they are used in motor tasks. These results might also be of interest for interaction designers. The results indicate that olfactory information is useful when planning and executing precise grasping tasks, which suggests that the representations elicited by olfactory information is highly detailed. The study seems to be well conducted and well grounded in previous literature. I have however some comments on the manuscript:


General comments

Outline of the manuscript: From reading the introduction and method, I find that the hypotheses and questions are somewhat scattered throughout the text. I find questions and hypotheses in both the introductory text, as well as in the method text (page 4, paragraph 2; p. 5, para 2; p.9; para 2 and on to page 10). Furthermore, the recap of the scientific question in the beginning of the discussion (p. 17) is somewhat different than the questions in the introduction. I think I understand what questions the authors want to answer, but it might become clearer if hypotheses and questions are clearly specified in one place.

Unclear and slightly too long method text: After several readings, I still do not fully grasp what statistical tests were used to answer the scientific questions. I understand the procedure clearly, but at the end of page 9, things begin to become unclear. I understand that the authors have used a planned orthogonal contrasts method, but is the plan to test the effects for "each participant at each epoch of the normalized movement time (from 10 to 100%, 10% steps) and for all the kinematic variables" (p. 10)? I am no expert on statistics, but as I remember, the number of planned comparisons should not be more than the available degrees of freedom.
At several places, the method gets a bit vague. At the end of page 10, the authors write "We then evaluated the degree to which the participants' sample exhibited the expected pattern of results by performing one-sample t-test." What is the expected pattern here, and how many t-tests were done? On page 11 the authors write "Here we used this procedure to examine possible differences in covariation patterns across experimental conditions..." What procedure is referred to here?
I think that a revision of the method (i.e. explanation of the statistical procedures used) can do great things in terms of accessibility and general understanding of the manuscript. I also believe that this section is a bit long. I suggest that the explanations of and arguments for specific statistical tests are removed.

Results: The results section suffers from some of the problems present in the methods section. The text is slightly too long, and I have difficulties understanding some figures. Especially Figure 10 and Figure 11 is difficult to follow. They depict color coded correlation coefficients for different points during the reaching, but the changes in correlations are difficult to follow. Can these results be presented in another fashion? My guess is that the length of the results section is caused by a very large data set. If the questions stated in the introduction can be answered even after omitting or merging some results in this large data set, it might be a revision to be taken into consideration.

Representations or cues? I have a general question of whether or not we can conclude that it is the actual representations elicited by the odours that forms the different grasping patterns, or if the subjects use the odours merely as cues for one of two motor actions. If the former case is true, we can conclude that odour representations are indeed rich in information. If the latter case is true, the subjects learn to associate an odorous stimulus (maybe even without accessing internal representations) to a specific course of action, not necessitating any semantic connection between odour and object. Would we for example get the same results even when exchanging the apple, strawberry, almond and orange odours for completely different odours? The authors describe the odours as "large" and "small", but these characteristics are taken from the objects, and not the actual odorants. I would much like a short comment on what it is the authors are studying. Is it "grasping representations" (p. 4, para 2), a "motor plan elicited by the olfactory stimulus" (p. 18, para 2) or indeed maybe even simple cues that initiates one of a few motor actions?

More specific comments
1. In the first paragraph of the introduction, the authors state that all sensory modalities are used in reach and grasp tasks. Is this correct? I find little use for example the taste system in these tasks.
2. On page 9, paragraph 1, the authors write "Abduction angles were reported on a continuum of negative values with respect to the baseline. An (sic!) decrease in such values indicated relatively greater abduction." Does a decrease in negative values imply lower values or higher values?

3. Letters have changed places? On page 8, three rows from the bottom, I suppose it should be "the fingers straightened, close (not closed) to each other"

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.