The Interval Censored Analysis menu is for fitting the Turnbull non-parametric maximum likelihood estimates to interval censored data using the TURNBULL procedure. Interval censored data occurs when the exact time point of an event cannot be observed, but you know a time before the event occurred (the Lower limit), and the time after the event occurred (the Upper limit). For example, if the subject in a trial was having weekly checkups, you might be able to know only that an event had occurred at some time between two checkups. Data may be right censored, where the event has not yet been observed before the end of the trial. A missing value should then be entered in the Upper limit variate. If an exact time of an event is known, this should be entered into both the lower and the upper limits. The analysis does not allow for left censored data, so values without the lower limit must be omitted from the data, as the procedure will fault missing lower limit values. If there are multiple groups in the data, these are specified as a factor in the Groups field.

## Available data

This lists data structures appropriate to the current input field. It lists either variates for use in specifying the lower and upper limits, or factors for specifying the groups. The contents will change as you move from one field to the next. Double-click a name to copy it to the current input field or type the name.

## Lower limit

Specify a variate containing the lower limits to be analysed. The lower limit is the last recorded time before the event occurred.

## Upper limit

Specify a variate containing the upper limits to be analysed. The upper limit is the first recorded time after the event occurred. If this variate contains missing values, these observations will be taken as right censored observations. The lower limits must always be less than or equal to the corresponding upper limit.

## Groups

An optional factor to specify the groups in the data. A separate model will be fitted for each group.

## Options

Opens the Options dialog for choosing which results to display and options controlling the fitting algorithm.

## Store

Opens the Store Options dialog for choosing to store the results from the analysis. This must be used prior to running the analysis.

## See also

- Options for choosing which results to display and fitting options
- Store options for choosing to save the results
- Kaplan-Meier (Interval based)
- Life-table menu for forming life-tables
- Nonparametric Survival Tests menu for nonparametric survival tests
- Survival Distributions menu
- Proportional Hazards menu
- TURNBULL procedure
- KAPLANMEIER procedure