On 2015 Dec 06, Gwinyai Masukume commented:

Unacknowledged important limitation - the effective sample size of 229 was insufficient to detect a statistically significant association.

Dean and colleagues find that there was a deficit of 229 pregnancies among women who were in the first trimester when the mass shooting at Port Arthur occurred and that by sex there was no statistically significant differential loss. They estimate that for every 100 females, 107 males were lost.

Although Dean and colleagues did not find a statistically significant difference by sex, their sample size of 229 was insufficient to detect the difference given the small effect size Austad SN, 2015.

A rigorous study Orzack SH, 2015 demonstrated that normally, in the absence of exogenous stressors such as that caused by the Port Arthur calamity, more female embryos/fetuses are lost during the first trimester, approximately 100 females for every 97 males.

In conclusion, the results presented by Dean and colleagues show excess male loss during a period when more female loss is expected. Thus, the small sample size of 229 should have been mentioned as an important limitation of the paper and a formal power calculation was apt.

On 2015 Dec 06, Gwinyai Masukume commented:

Unacknowledged important limitation - the effective sample size of 229 was insufficient to detect a statistically significant association.

Dean and colleagues find that there was a deficit of 229 pregnancies among women who were in the first trimester when the mass shooting at Port Arthur occurred and that by sex there was no statistically significant differential loss. They estimate that for every 100 females, 107 males were lost.

Although Dean and colleagues did not find a statistically significant difference by sex, their sample size of 229 was insufficient to detect the difference given the small effect size Austad SN, 2015.

A rigorous study Orzack SH, 2015 demonstrated that normally, in the absence of exogenous stressors such as that caused by the Port Arthur calamity, more female embryos/fetuses are lost during the first trimester, approximately 100 females for every 97 males.

In conclusion, the results presented by Dean and colleagues show excess male loss during a period when more female loss is expected. Thus, the small sample size of 229 should have been mentioned as an important limitation of the paper and a formal power calculation was apt.