# How do I merge two age groups per 10k from an abortion rate table?

#### samlag

##### New member
• I would like to find the 12 to 19 abortion rate per 10,000 women from the following table:

This would be the merging of age groups "12-17" with "18-19" to some how get the 12-19 abortion rate.

The answer I came up with is 125.67 per 10,000. Not sure if that is correct.
I multiplied the rate per individual woman by the number of abortions for each group which I think gave the number of women. I added both of these groups together. I added both groups number of abortions and then divided the amount of women by amount of abortions and multiplied it by 10,000.

#### tkhunny

##### Moderator
Staff member
You need the Total Number of each population or you cannot do it. Your value is much too high.

For example - from baseball:

Given a batting average of 0.250 in April and a batting average of 0.290 in May, there are lots of wrong ways to calculate the batting average for both months, combined.

(0.250 + 0.290)/2 = 0.270 = Wrong!

Hits/At-Bats: April: 25/100, May: 33/114 ==> Average for both: (25+33)/(100+114) = 58/214 = 0.271 -- In this case, just a little higher than the bad way.

Note: In this tiny example, it made very little difference. Don't begin to believe this is always the case. We're just showing that it IS different. Depending on your data, it can be VERY different. Also, if our baseball player has a bonus in his contract for hitting OVER 0.270, he will appreciate the correct calculation. A very tiny difference can mean quite a lot.

1) We are given the Number of Abortions.
2) We are NOT given the numbers of the total populations. If we can't figure that out, we're sunk. We can say something like, "between 0 and 240"

How do you propose to reconstruct the total population? Think about that incidence number. How is it calculated?

Note: "rate per individual woman" doesn't mean anything. Most of those are zero (0) over probably any period, long or short.

Last edited:

#### samlag

##### New member
There appears to be a way to reverse engineer the data provided to estimate the populations,

Also to be noted, "Distribution of abortion age" was translated from "Répartition des ges l'avortement" in French. It appears to me to be the percentage each age group's abortions are of the total amount of abortions.