By Chris Campbell
As another new baby card is passed around the office, and the latest cute baby pictures are emailed out, a discussion is underway. Could it be true? Something in the water? An elixir of fertility that should be bottled and sold to desperate couples for enormous profit? Is Mango having some crazy baby boom?!
Mango has been having something of a population explosion. In the first seven years of Mango there were only a couple of babies. Since moving to Methuen Park it seems like there’s been one every few months. So is there really something magical happening here?
Of course not. Now demonstrating this kind of data is an absolute HR nightmare, so please forgive me if I’m somewhat coy with revealing the raw datasets. However, I’d like to share a small analysis I did as an example of how to set up a dataset for survival analysis with multiple events. To investigate the hypothesis that there’d been a change in the rate of baby production, I had two datasets. One, babies, has the Date of birth of the child, as well as some covariates. Splitting births into events in the first 3432 days (9.4 years) and most recent 1268 days (3.5 years) shows that 40% the number of babies were born before moving to our current home, Methuen Park, compared to when at the new offices. In fact, the rate jumps from 0.4 babies/year to 2.9 babies/year!
names(babies)
# [1] "Name" "Date" "Sex" "Parent"
library(dplyr)
day1 3432),
Number = length(Date))
# Source: local data frame [2 x 2]
#
# Methuen Number
# (lgl) (int)
# 1 FALSE 4
# 2 TRUE 10
I helped …read more
Source:: r-bloggers.com
