- 104 Views
- Uploaded on
- Presentation posted in: General

What is Event History Analysis?

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

What is Event History Analysis?

Fiona Steele

Centre for Multilevel Modelling

University of Bristol

- Example applications of EHA
- Event history data and possible sources
- Methods of analysis with application to timing of 1st partnership
- Descriptive analysis
- Event history modelling

- Further topics

Methods for analysis of length of time until the occurrence of some event. The dependent variable is the duration until event occurrence.

- EHA also known as:
- Survival analysis (particularly in biostatistics and when event is not repeatable)
- Duration analysis
- Hazard modelling

- Education – time to leaving full-time education (from end of compulsory education); time to exit from teaching profession
- Economics – duration of an episode of unemployment or employment
- Demography – time to first birth (from when?); time to first marriage; time to divorce
- Psychology – duration to response to some stimulus

Types of Event History Data

- Dates of start of exposure period and events, e.g. dates of start and end of an employment spell
- Usually collected retrospectively
- UK sources include BHPS and cohort studies (partnership, birth, employment, and housing histories)

- Current status data from panel study, e.g. current employment status each year
- Collected prospectively

- Durations are always positive and their distribution is often skewed
- Censoring – there are usually people who have not yet experienced the event when we observe them
- Time-varying covariates – the values of some covariates may change over time

Censoring

- Right-censoring is the most common form of censoring. Durations are right-censored if the event has not occurred by the end of the observation period.
- E.g. in a study of divorce, most respondents will still be married when last observed

- Excluding right-censored observations leads to bias and may drastically reduce sample size
- Usually assume censoring is non-informative

Key Quantities in EHA

- In EHA, interest is usually focused on the hazard functionh(t) and the survivor function S(t)
- h(t) is the probability of having at event at time t, given that the event has not occurred before t
- S(t) is the probability that an event has not occurred before time t

- Group durations into intervals t=1,2,3,… (often already in this form)
- Record no. at risk at start of interval r(t), no. events during interval d(t), and no. censored during interval w(t)
- An estimate of the hazard is d(t)/r(t). Sometimes there is a correction for censoring

Estimator of survivor function for interval t is

E.g. probability of surviving to the start of 3rd interval

= probability no event in 1st interval and no event in 2nd interval

Source: Subsample from the National Child Development Study

- h(16)=0.02 so 2% partnered at age 16
- h(20)=0.13 so of those who were unpartnered at their 20th birthday, 13% partnered before age 21
- S(20)=0.77 so 77% had not partnered by age 20

Introducing Covariates: Event History Modelling

There are many different types of event history model, which

vary according to:

- Assumptions about the shape of the hazard function
- Whether time is treated as continuous or discrete
- Whether the effects of covariates can be assumed constant over time (proportional hazards)

The Cox Proportional Hazards Model

The most commonly applied model which:

- Makes no assumptions about the shape of the hazard function
- Treats time as a continuous variable
- Assumes that the effects of covariates are constant over time (although this can be modified)

hi(t) is hazard for individual i at time t

xi is a covariate with coefficient β

h0(t) is the baseline hazard, i.e. hazard when xi=0

The Cox model can be written

hi(t) = h0(t) exp(βxi)

or sometimes as

log hi(t) = log h0(t) + βxi

Note: x could be time-varying, i.e. xi(t)

- exp(β) - also written as eβ - is called the relative risk
- For each 1-unit increase in x the hazard is multiplied by exp(β)
- exp(β)>1 implies a positive effect on hazard, i.e. higher values of x associated with shorter durations
- exp(β)<1 implies a negative effect on hazard, i.e. higher values of x associated with longer durations

The hazard of partnering at age t is 1.5 times higher for women

than for men.

So women partner at an earlier age than men.

We assume that the gender difference in the hazard is the same for

all ages.

- Event times are often measured in discrete units of time, e.g. months or years, especially when collected retrospectively
- Before fitting a discrete-time model we must restructure the data so that we have a record for each time interval

The response variable for a discrete-time model is the binary indicator of event occurrence yi(t).

The hazard function is the probability that yi(t)=1.

Fit a logistic regression model of the form:

- FEMALERespondent’s sex (1=female, 0=male)
- FULLTIME(t)Whether in full-time education at age t
- (1=yes, 0=no)
- We need to choose the form of α(t). Try:
- step function (dummy variable for each year age 16-33)
- quadratic function by including t and t2
- Choice will be guided by plot of the hazard function.

Reference

category for t

=age 16

Exp(B) are (multiplicative) effects on the odds of partnering at age t

Women partner more quickly than men.

Enrolment in full-time education is associated with a

(1-0.324)100=68% reduction in the odds of partnering, i.e. a delay

- So far we have assumed that the effects of x are the same for all values of t
- It is straightforward to relax this assumption in a discrete-time model by including interactions between x and t in the model
- The following graphs show the predicted log-odds of partnering from 2 different models: 1) the ‘main effects’ model on the previous slide, 2) a model with interactions t*female and t2*female added.

Proportional Gender Effects

- Repeated events, e.g. multiple marriages or births
- Multilevel modelling

- Competing risks, e.g. different reasons for leaving a job (switch to another job, redundancy, sacked)
- Fit set of logistic regression models or a single multinomial model

- Multiple states, e.g. may wish to model transitions between unpartnered, marriage and cohabitation states
- Include dummies for state, and interact with duration and covariates

- Multiple processes, e.g. joint modelling of partnership and education histories

Singer, J.D. and Willet, J.B. (1993) “It’s about time: Using discrete-time

survival analysis to study duration and the timing of events.” Journal

of Educational Statistics, 18: 155-195.

Blossfeld, H.-P. and Rohwer, G. (2007) Event History Analysis with Stata.

Mahwah (NJ): Lawrence Erlbaum.

Steele, F. (2005) Event History Analysis. NCRM Review Paper NCRM/004.

Downloadable from http://www.ncrm.ac.uk/publications/index.php.

Steele, F., Goldstein, H. and Browne, W. (2004) “A general multistate

competing risks model for event history data, with an application to a study of

contraceptive use dynamics.” Statistical Modelling, 4: 145-159.