Regression table – probit model for Probit Analysis

Probit regression examines the relationship between two variables:
  • A binomial response, which is whether the unit survived or failed after the stress was imposed.
  • A continuous stress variable, which is the measurement of the stress imposed on the unit.

The probit model uses the selected cumulative distribution function to link the level of stress to a probability of failure. Use the probit model to examine how the probability of an event changes as the level of stress changes and to predict the probability of an event for any stress value in the experimental range.

The Constant coefficient is the value of the inverse cumulative distribution function when no stress is imposed and the natural response rate is 0. The natural response rate is the probability that a unit fails without being exposed to any of the stress. This statistic is used in situations with high mortality or high failure rates. If the natural response rate is greater than 0, then the stress does not cause all of the failures in the analysis.

Example output

Regression Table Standard Variable Coef Error Z P Constant -6.20376 1.06565 -5.82 0.000 Stress 0.0089596 0.0015615 5.74 0.000 Natural Response 0 Log-Likelihood = -38.516


For the windshield data, the estimate of the Stress coefficient, β1, is 0.0089596 and the estimate of the Constant coefficient, β0, is −6.20376. The positive coefficient for Stress indicates that increasing velocity increases the probability of windshield breakage.

By using this site you agree to the use of cookies for analytics and personalized content.  Read our policy