Annals of Saudi Medicine: RE: Perceptions of confidence among Saudi nursing interns during internship program: a cross-sectional study The use of nonparametric statistical tests
RE: Perceptions of confidence among Saudi nursing interns during internship
program: a cross-sectional study The use of nonparametric statistical tests
================================================================================
Mustafa Afifi on 06/12/2018 10:00:00
To The Editor: I have read with care the paper of Aboshaiqah et at. entitled
“Perceptions of confidence among Saudi nursing interns during internship
program: a cross-section.al study”.1 I acknowledge the output of the authors,
but found statistical errors which I would kindly like to bring to their notice.
Despite that the indicators in their research showed evidence of normal
distribution of their data and linearity of variables, the authors used
nonparametric statistical tests (Mann Whitney U test and Spearman Rank
correlation). Actually, their sample size is big enough (n=301 interns). Also,
the mean value of the overall perceived confidence score equals the me.dian and
has a standard deviation (SD) less than one third of this mean (both the mean
and median=3.7, SD 0.9). Additionally, the mean value of the overall perceived
impotence score almost equals the median and has a SD less than one third of the
mean (the mean=4.1, SD 0.9, median=4.4). The median should appear as a integer
and not as a decimal. There are two variables in Table 5 which were not used in
Table 4.
Besides, the figures of perceived confidence and impotence were presented in the
abstract in mean values and the authors wrote only the P of significance for
Spearmen correlation and only for the confidence part of the SECS. I think the
authors should clarify in their methods why they used the nonparametric
statistics especially when para.metric tests are more robust in this case.
Kitchen stated in her paper that the assumptions for non-parametric tests are
weaker than those of parametric and researchers are better only resort to the
first if normality of data was not met or the underlying distributions are
heavily tailed or extremely skewed and/or the sample size is small (n
References
1. Ahmad E. Aboshaiqah, Regie B. Tumala, Joel G. Patalagsa, Ahmed A.
Al-Khaibary, Haya Al Fozan, John Paul Ben T. Silang. PerCeptions of confidence
among Saudi nursing interns during internship program: a cross-sectional study.
Ann Saudi Med. 2018 Jul-Aug; 38(4): 288–298.
2. Kitchen, Christina M. R. Nonparametric versus parametric tests of location in
biomedical research. Am J Ophthalmol. 2009 Apr; 147(4): 571–572.