reviewers
- This paper presents an AR-based CAD system which allows gesture-based solid modelling. It presents two use cases, and a (supposed) user study.
I think that the topic of the paper is quite relevant to the intended audience, and in this sense it is APPropriate for this journal.
However, the information presented in the paper is far from being acceptable. First, the english writing is unacceptable, making some paragraphs unintelligible. The introduction is not readable, making the reader miss the motivation and/or the problem that this work tries to solve (beyond designing yet another AR-Based CAD system). It neither encourages readers to go on reading. The abstract does not describe in an accurate manner the purpose of the paper or its contribution. Instead, it enumerates (more or less) the sections of the paper. The rest of the sections also are hardly readable, due to the misuse of punctuation marks, and the organization of the paper does not help in the understanding of the approach.
My main concern, overall, is the technical soundness of the paper. First, the proposed AR system design does not include any new algorithm/method/procedure except the fact of applying gesture-based solid modelling (a known technique) to a CAD system.
Second (and most important), the evaluation of its impact (user studies) is very poor (see below). Thus, it does not represent a practical contribution to this field of research.
The measurement of the impact of new technologies in human users requires a solid procedure and statistical study, like the one shown in [1]. moreover, and analysis of the collected data to find if they follow or not a normal distribution should be made. The Kolmogorov-Smirnov tests [3], Anderson-Darling tests [2], as well as Shapiro-Wilk tests [4] should be applied to collected data. In case these data do not follow a normal distribution, non-parametric tests like Wilcoxon-Mann-Whitney [5] for unpaired data should be used. Otherwise, parametric tests (the t-test and the Cohens test for paired data and the ANOVA test) can be used. Another crucial point is that experiments in human-computer interaction requires a Minimum population size of 40 people, as shown in [6,7,8].
[1] Arino, J., Gil-Gómez, J., Lizandra, M.D., & Vayá, R.M. (2014). A comparative study using an autostereoscopic display with augmented and virtual reality. Behaviour & IT, 33, 646-655.
[2] T. W. Anderson. Anderson–Darling Tests of Goodness-of-Fit, pp. 52–54.
Springer Berlin Heidelberg, Berlin, Heidelberg, 2011. doi: 10.1007/978-3
-642-04898-2 118
[3] F. J. Massey. The Kolmogorov-Smirnov test for goodness of fit. Journal
of the American Statistical association, 46(253):68–78, 1951.
[4] S. S. Shapiro and M. B. Wilk. An Analysis of Variance Test for Normality
(Complete Samples). Biometrika, 52(3/4):591–611, 1965.
[5] M. Neuhäuser. Wilcoxon–Mann–Whitney Test, pp. 1656–1658. Springer
Berlin Heidelberg, 2011. doi: 10.1007/978-3-642-04898-2 615
[6] K. Hornbaek. Some whys and hows of experiments in human computer
interaction. Foundations and Trends in HumanComputer Interaction,
5(4):299–373, 2013. doi: 10.1561/1100000043
[7] D. C. Montgomery and G. C. Runger. Applied Statistics and Probability
for Engineers. John Wiley and Sons, 2003.
[8] J. M. Morse. Determining sample size. Qualitative Health Research,
10(1):3–5, 2000. doi: 10.1177/104973200129118183