During any given pregnancy a woman has the exact same chance of giving birth to a boy as she does a girl. Since you’re asking about birth *order* rather than the total number of boys and girls a woman gives birth to, we have to look at each individual pregnancy.

Pregnancy #1: 50% chance for a boy, 50% chance for a girl.
Had a boy.

Pregnancy #2: 50% chance for a boy, 50% chance for a girl.
Had a boy.

Pregnancy #3: 50% chance for a boy, 50% chance for a girl.
Had a boy.

Pregnancy #4: 50% chance for a boy, 50% chance for a girl.
Had a girl.

Since we’re looking at each pregnancy individually, we cannot say that she’s more likely to have a girl than a boy simply because she’s already had ‘x’ number of boys.

Now we CAN say that OVERALL, she should have 50% male children and 50% female children. The woman actually has 75% male children and 25% female. Why? Small sample size. Just as it wouldn’t be odd to have flipped 3 heads and only 1 tails in four coin flips, it wouldn’t be odd to have 3 sons and only 1 daughter in four pregnancies. If the woman were to have 100 children where 75 were male and only 25 were female, then that would indeed be more than just a coincidence chalked up to sample size.
References :