Survival Analysis Theory

Property 1:

S(t) = eH(t)

or equivalently

H(t) = −ln S(t)

Proof: Since S(t) = 1 − F(t), the derivative S′(t) = − F′(t) = − f(t).  Thus

image106x

Taking the integral of both sides yields

image107x

From which we get H(t) = -ln S(t). Taking the exponential of both sides yields the first result.

Observation: Some properties of the hazard function are

image059ximage060x

Observation: If the hazard function is constant then t has an exponential distribution; i.e. a constant hazard function corresponds to an exponentially distributed lifetime. To see this, let

image061

Then c(1−F(t)) = F′(t), which has a solution when F(t) = 1 − ce-ct. Alternatively

image062x

Reference

Wikipedia (2015) Survival analysis
https://en.wikipedia.org/wiki/Survival_analysis#:~:text=Survival%20analysis%20is%20a%20branch,and%20failure%20in%20mechanical%20systems.

6 thoughts on “Survival Analysis Theory”

  1. Good day. I am a PhD student in Nigeria researching on a topic ‘Information technology skills, demographic and institutional factors as correlates of doctoral degree completion in Nigerian Library Schools. I used survival analysis theory as the theoretical framework. I want to find out if this theory can actually be used for this study.
    Thank you as I look forward to your quick response.

    Reply
    • Yusuf,
      Whether it is appropriate for your specific study, I can’t say, but, in general, survival analysis can be used in studies about completion of a school degree.
      Charles

      Reply

Leave a Comment