On 2015 Mar 25, Janet Kern commented:

There is a statistical error in this research study. This error can be seen in Table 2 at 24 months of age. By utilizing the numbers provided in Table 2 (see below) it is evident that the difference between cases and controls at 24 months is highly statistically significant. The journal, Vaccine, was notified of the error. However, since, to date, no clarification has been issued, it is important to note that the conclusions seen in the abstract above are misleading and are the opposite of the conclusion supported by the data. The corrected results indicate that there is a statistically significant relationship between Thimerosal exposure and autism spectrum disorder.

***** At 24 months from the data provided using a t-test reveals the following:

Unpaired t test Mean of * sample 1 from summary = 804.2 (n = 189) Mean of * sample 2 from summary = 632.1 (n = 224)

Assuming equal variances Combined standard error = 71.838701 df = 411 t = 2.395645 One sided P = 0.0085 Two sided P = 0.017 95% confidence interval for difference between means = 30.882882 to 313.317118 Power (for 5% significance) = 90.07%

Assuming unequal variances Combined standard error = 72.061016 df = 394.166765 t(d) = 2.388254 One sided P = 0.0087 Two sided P = 0.0174 95% confidence interval for difference between means = 30.445864 to 313.754136 Power (for 5% significance) = 66.35%

Update 5/24/2015: When the journal Vaccine was notified of the error, it notified the authors. In response to the notification of the error, the authors changed the numbers in Table 2 of their study. The authors changed the mean and standard deviation for the controls at 24 months from 632.1 (715.1) to 676.8 (719.5). No explanation for the error or justification for the change was given.

To date, the journal Vaccine and the study authors have refused to release the study dataset for further evaluation.

On 2015 Mar 25, Janet Kern commented:

There is a statistical error in this research study. This error can be seen in Table 2 at 24 months of age. By utilizing the numbers provided in Table 2 (see below) it is evident that the difference between cases and controls at 24 months is highly statistically significant. The journal, Vaccine, was notified of the error. However, since, to date, no clarification has been issued, it is important to note that the conclusions seen in the abstract above are misleading and are the opposite of the conclusion supported by the data. The corrected results indicate that there is a statistically significant relationship between Thimerosal exposure and autism spectrum disorder.

***** At 24 months from the data provided using a t-test reveals the following:

Unpaired t test Mean of * sample 1 from summary = 804.2 (n = 189) Mean of * sample 2 from summary = 632.1 (n = 224)

Assuming equal variances Combined standard error = 71.838701 df = 411 t = 2.395645 One sided P = 0.0085 Two sided P = 0.017 95% confidence interval for difference between means = 30.882882 to 313.317118 Power (for 5% significance) = 90.07%

Assuming unequal variances Combined standard error = 72.061016 df = 394.166765 t(d) = 2.388254 One sided P = 0.0087 Two sided P = 0.0174 95% confidence interval for difference between means = 30.445864 to 313.754136 Power (for 5% significance) = 66.35%

Update 5/24/2015: When the journal Vaccine was notified of the error, it notified the authors. In response to the notification of the error, the authors changed the numbers in Table 2 of their study. The authors changed the mean and standard deviation for the controls at 24 months from 632.1 (715.1) to 676.8 (719.5). No explanation for the error or justification for the change was given.

To date, the journal Vaccine and the study authors have refused to release the study dataset for further evaluation.