The null hypothesis is that the treatment does not affect the likelihood of the baby being born a girl.
The alternative is that the treatment increases the likelihood of the child being born a girl.
If the null and alternative hypothesis were not clear, start with the alternative hypothesis. The alternative hypothesis is the thing or effect that the study is meant to find, if it exists. To get the null hypothesis, take the alternative hypothesis and replace "increase" or "decrease" with "has no effect", replace "greater than" or "less than" with "equal to". It is the boring conclusion that you get if you find nothing.
First blank: Is not statistically significant. "There is about a 16% chance of getting that many girls if the method had no effect". In other words, even if the null hypothesis were true, we had a 16% chance of finding 1031 or more girls in our sample of births given the sample size. That means the p-value is 16%, since the p-value is the probability of finding that estimate or an estimate farther in the direction of the alternative hypothesis if the null hypothesis is true. A p-value greater than 5% is not statistically significant because the null hypothesis explains the data well.
Second blank: Not many couples would likely use a procedure that raises the likelihood of a girl from the approximately 50% to
Third blank: To the 51.1% produced by this method. The percentage girls in the sample of births was