3.8 Calculate 95% CIs

Run the following code to calculate the 95% confidence interval for the Father_age slope parameter.

confint(fit_pat, 'Father_age', level = 0.95)

##               2.5 %  97.5 %
## Father_age 1.243118 1.45386

So, 95% of the time, the number of additional DNMs per year of paternal age is between 1.24 and 1.45.

Modify your code to get the 95% CI for the Mother_age slope. What’s the interpretation of this confidence interval?

confint(fit_mat, 'Mother_age', level = 0.95)

##                2.5 %    97.5 %
## Mother_age 0.3094713 0.4474528

95% of the time, the number of additional DNMs per year of maternal age is between 0.31 and 0.45.

Now that we have the confidence intervals for both slope parameters, we can finally compare them.

Our two CI ranges are non-overlapping. The paternal range is [1.24, 1.45] and the maternal range is [0.31, 0.45].

If the 95% CIs for two parameters don’t overlap, this strongly supports that the parameters are significantly different from one another. So, it seems likely that paternal and maternal gametes experience different rates of de novo mutation.

If the CIs for two parameters overlap, are they not significantly different?

Not necessarily. More analysis, like a hypothesis test, is needed to make a final decision.