Non-parametric Survival Analysis in JASP

Survival analysis is an ubiquitous tool primarily used for evaluating interventions in clinical trials and device assessments in medicine, but it is also commonly used for assessing reliability of components in engineering etc. It allows us to analyze time-to-event data in case some of the observations are censored–i.e., not all events occurred in the follow-up timeframe and we know that an event would have happened later on.

In this blogpost, we reproduce several outputs from Chapter 4: “Non-parametric methods: the Kaplan-Meier estimator” in Melinda Mill’s (2011) “Introducing Survival and Event History Analysis” textbook. This is our first, and long awaited, step in adding survival analysis to JASP. The upcoming releases will expand the Survival module with further functionality (e.g., semi-parametric, parametric, Bayesian,…).

Example Data

We use the “Leukemia” data set which provides survival times in patients with Acute Myelogenous Leukemia. The data were collected in order to assess whether the standard course of chemotherapy should be extended (‘maintainance’) for additional cycles. The data set contains the following variables:

time – Survival time.

status – Whether time corresponds to an event, death, (status = 1) or a censored observation (status = 0).

x – Whether the standard course of chemoterapy was extended for additional cycles (x = Maintained) or not (x = Nonmaintained).

(The data set and description was obtained from the R package ‘survival’.)


We reproduce several outputs Mill’s (2011) textbook: Kaplan-Meier survival curves estimate which is automatically produced once the Time to Event, Event Status, and Event Indicator variables are selected, tests for difference in survival curves between strata defined using the Factors variables and the Tests checkboxes, visualization of the survival curves using the Survival curve plot checkboxes, and life table summary using the Life table checkbox. Other options, like visualization of the risk table, cumulative events table, and censoring plot are available in the menu.


The Kaplan-Meier Summary Table reproduces the R output on page 75. (Note that the empty cells correspond to NA values.)

The Tests Table reproduces the first, fourth, and fifth row of Table 4.5. The sixth row can be reproduced by setting ‘Rho = 1’ under the ‘Tests’ option.

The Survival Curve figure reproduces the Figure 4.5 with addition of confidence intervals.

The ‘Life Table’ reproduces the R output on pages 77-78.


In this blogpost we reproduced several key outputs of non-parametric survival analysis from Mills’ (2010) textbook.


Mills, M. (2010). Introducing Survival and Event History Analysis. SAGE.

Therneau, T. (2023). A Package for Survival Analysis in R. R package version 3.5-7,

About the author

František Bartoš

František Bartoš is a PhD candidate at the Psychological Methods Group of the University of Amsterdam.