Responses
Other responses
Jump to comment:
- Published on: 23 November 2015
- Published on: 23 November 2015
- Published on: 23 November 2015
- Published on: 23 November 2015Prognosis of patients with silent versus symptomatic myocardial ischemiaShow More
Dear Editor,
We do appreciate Dr Tsakok's concern regarding the correct use of the two-sample-t-test. At our centre it is common use to apply the Kolmogorv-Smirnov test to test the normality assumption in all our analyses. In our study this test was not significant and therefore, we were allowed to use the parametric tests such as the two sample t-test. The method Dr Tsakok is proposing is unknown to us. We are not...
Conflict of Interest:
None declared. - Published on: 23 November 2015Comment on heart analysesShow More
Dear Editor,
Biagini et al (1), Rodriguez et al (2) and Dursunglu et al (3) illustrate incorrect statistical analyses, by using the two-sample t-test. This compares normal means assuming unknown but equal variances. While the Central Limit Theorem justifies normality for mean inferences, making it unnecessary to test (3) for normality, unknown variances need not be homogeneous. This makes the two-sample t-test un...
Conflict of Interest:
None declared. - Published on: 23 November 2015More work on this neededShow More
Dear Editor,
To suggest that “patients with asymptomatic ischaemia should be treated with a complete medical therapy or revascularisation as patients with symptomatic myocardial ischaemia” is a little too strong a recommendation to make from an observational study. The topic of your paper is of vital interest, but surely a randomised-controlled trial is needed with clinically important end-points? Potential confou...
Conflict of Interest:
None declared.