Trying to figure out the probability of something.

[nothingkwt]

Say I get married, the odds of me getting a divorce in a year or less are 11.5% and the chances for divorce for marriages between 1-5 years is 15.7%. My first question is, would adding them up be a good approximation for the odds of divorce in 5 years? Bearing in mind that the total number of marriages and divorces are 9597 and 4261 respectively.

My main question though is, if the odds of divorce for husbands in the age group of 25-29 is around 27%. And for the group of 30-34 the divorce rate is ~21%.

How can you calculate the divorce rate for someone who married at 28 while also taking into account the length of the marriage? I feel like my question isn't clear but I'm not sure how to explain it better than this.

 

[HallsofIvy]

It's a little more complicated than that. The probability of divorce in the first year is 0.115 and the probability of divorce in years 1 to 5 is 0.157. But, in order to get divorced in "years 1 to 5" you must have NOT gotten divorced in the first year- and that is 1- 0.115= 0.885, So the probability of getting divorce in year 5 or earlier is 0.115+ (0.885)(0.157)

 

[nothingkwt]

It's a little more complicated than that. The probability of divorce in the first year is 0.115 and the probability of divorce in years 1 to 5 is 0.157. But, in order to get divorced in "years 1 to 5" you must have NOT gotten divorced in the first year- and that is 1- 0.115= 0.885, So the probability of getting divorce in year 5 or earlier is 0.115+ (0.885)(0.157)

Yeah, that makes sense. But what about the other question?

 

[HallsofIvy]

You start by saying "the odds of me getting a divorce in a year or less are 11.5% and the chances for divorce for marriages between 1-5 years is 15.7%." but in your second question say "the odds of divorce for husbands in the age group of 25-29 is around 27%. And for the group of 30-34 the divorce rate is ~21%." Are those for the first year as you give originally or just generic? If the 27% and 21% are for getting divorced ever, then you cannot use them to calculate "year per year".

 

[jbriggs444]

My main question though is, if the odds of divorce for husbands in the age group of 25-29 is around 27%. And for the group of 30-34 the divorce rate is ~21%.

How can you calculate the divorce rate for someone who married at 28 while also taking into account the length of the marriage? I feel like my question isn't clear but I'm not sure how to explain it better than this.

From the information provided, there is no way to know. Perhaps people who marry at 28 never ever get divorced. Perhaps people who marry at 28 all get divorced the day after the wedding. Perhaps only one person in the world ever got married at age 28. All three possibilities are consistent with the information that has been given.

What you could do is come up with a plausible model of the complete continuous joint distribution of the probability of divorce after x years for a person who married at age y. As above, there are many (infinitely many) such models that fit the given's of the situation.

You might possibly come up with a reasonable family of models where the family members vary based on a set of parameters. Then you could select the parameter values that are a "best fit" to the givens of the problem.

 

[nothingkwt]

You start by saying "the odds of me getting a divorce in a year or less are 11.5% and the chances for divorce for marriages between 1-5 years is 15.7%." but in your second question say "the odds of divorce for husbands in the age group of 25-29 is around 27%. And for the group of 30-34 the divorce rate is ~21%." Are those for the first year as you give originally or just generic? If the 27% and 21% are for getting divorced ever, then you cannot use them to calculate "year per year".

No those percentages are generic and not tied to any duration. I don't understand what year per year means?

 

[nothingkwt]

From the information provided, there is no way to know. Perhaps people who marry at 28 never ever get divorced. Perhaps people who marry at 28 all get divorced the day after the wedding. Perhaps only one person in the world ever got married at age 28. All three possibilities are consistent with the information that has been given.

What you could do is come up with a plausible model of the complete continuous joint distribution of the probability of divorce after x years for a person who married at age y. As above, there are many (infinitely many) such models that fit the given's of the situation.

You might possibly come up with a reasonable family of models where the family members vary based on a set of parameters. Then you could select the parameter values that are a "best fit" to the givens of the problem.

These numbers are actually not made up but are from a statistic here where I live, there's actually a lot more information on the website. The statistic only gave the number of marriages and divorces and I calculated those percentages. I think I understand what you mean by the model and I'll try to make it.