Select the options for Fit Cox Model in a Counting Process Form

Stat > Reliability/Survival > Cox Regression > Fit Cox Model in a Counting Process Form > Options

Method for handling tied event times

Specify the method Minitab uses to handle ties. Usually the Efron method provides better estimates than the Breslow method when there are many ties in the response data. The two methods yield the same estimates when there are no ties in the response data.

Confidence level for all intervals

Enter the level of confidence for the confidence intervals for the coefficients, relative risks, and survival function.

Usually, a confidence level of 95% works well. A 95% confidence level indicates that, if you took 100 random samples from the population, the confidence intervals for approximately 95 of the samples would contain the mean response. For a given set of data, a lower confidence level produces a narrower interval, and a higher confidence level produces a wider interval.

Type of confidence interval

You can select a two-sided interval or a one-sided bound. For the same confidence level, a bound is closer to the point estimate than the interval. The upper bound does not provide a likely lower value. The lower bound does not provide a likely upper value.

For example, the predicted mean concentration of dissolved solids in water is 13.2 mg/L. The 95% confidence interval for the mean of multiple future observations is 12.8 mg/L to 13.6 mg/L. The 95% upper bound for the mean of multiple future observations is 13.5 mg/L, which is more precise because the bound is closer to the predicted mean.
Two-sided
Use a two-sided confidence interval to estimate both likely lower and upper values for the unknown parameter.
Lower bound
Use a lower confidence bound to estimate a likely lower value for the unknown parameter.
Upper bound
Use an upper confidence bound to estimate a likely higher value for the unknown parameter.

Variance-covariance matrix for analysis

From the drop-down list, select Robust variance-covariance to perform the analysis using the robust covariance matrix1 for the parameter estimates. When you select this option, all the tests and confidence intervals in the analysis use the robust covariance matrix.

You can specify a column in Cluster identification for robust covariance matrix (optional) to identify groups of correlated observations due to the study design. Rows with the same value are clustered observations. For example, in recurrent event models where each subject can experience the event multiple times, the observations within the same subjects are correlated. If you specify a column, Minitab calculates the robust covariance to account for the presence of clustered observations. If you do not specify a column, the effect is the same as if you use a column with a different value in every row.

The input column can be numeric, text or date/time. Minitab includes missing values when it calculates the robust variance-covariance and groups them together in the analysis.

Test for ANOVA table

Specify the test Minitab uses for the ANOVA table. Empirical studies have shown that the convergence rates of the Likelihood ratio test and Wald test are similar. The Score test converges less rapidly to the limiting chi-squared distribution.

When you select Robust variance-covariance in the Variance-covariance matrix for analysis drop-down list, the ANOVA table always displays the Wald test because the Likelihood ratio test and Score test assume that the observations within clusters are independent.

Case identification (for subject residuals)

Specify a column to identify the subjects in the study. The column can be numeric, text, or date/time. The column must be the same length as the start time or the end time column. You can identify a subject using a case number, ID, or name. Minitab groups missing rows together to calculate the per case diagnostics.

In a counting process input format, a single subject can have multiple rows of diagnostic statistics. If you specify a case-identification column, Minitab provides a single diagnostic statistic per case by summing the statistics over the multiple rows pertaining to the same case.

1 Lin, D.Y., and Wei, L.J. (1989). The robust inference for the Cox Proportional hazards model. Journal of the American Statistical Association, 84: 1074-1078