Problem: Data set is given (Data-B.
csv) for the information of a particular firm’s
employees. Compute the Kaplan-Meier estimates of the survival function and write
a report based on your findings.
Methodology:
Suppose there are observations on 𝑛 individuals and there are 𝑘 (𝑘 ≤ 𝑛) distinct
times 𝑡1 < 𝑡2 < ⋯⋯ < 𝑡𝑘 at which deaths occurs. Let 𝑑𝑗 be the number of deaths
at time 𝑡𝑗. Let 𝐿𝑖 be the censoring times for individuals whose lifetimes are not
observed. Then Kaplan-Meier estimate of 𝑆 𝑡 is defined as
nj−dj
S ( t )= ∏
nj
Where 𝑛𝑗 is the number of individuals at risk at time 𝑡𝑗 i.e. the number of
j : tj<t
individuals alive and uncensored just prior 𝑡𝑗.
Results:
time [Link] [Link] survival [Link] lower 95% CI upper 95% CI
73 701 12 0.9829 0.00490 0.9733 0.9925
110 682 1 0.9814 0.00510 0.9715 0.9915
146 681 3 0.9771 0.00566 0.9661 0.9883
183 677 4 0.9713 0.00632 0.9590 0.9838
219 672 3 0.9670 0.00677 0.9538 0.9804
256 667 4 0.9612 0.00732 0.9470 0.9757
292 659 1 0.9597 0.00745 0.9453 0.9745
329 655 1 0.9583 0.00758 0.9435 0.9733
365 652 6 0.9495 0.00832 0.9333 0.9659
402 641 7 0.9391 0.00911 0.9214 0.9571
438 633 8 0.9272 0.00991 0.9080 0.9469
475 616 4 0.9212 0.01030 0.9012 0.9416
511 610 8 0.9091 0.01101 0.8878 0.9310
548 601 13 0.8895 0.01205 0.8662 0.9134
584 584 11 0.8727 0.01284 0.8479 0.8982
621 566 11 0.8557 0.01357 0.8296 0.8828
657 549 12 0.8370 0.01431 0.8095 0.8656
694 532 16 0.8119 0.01520 0.7826 0.8422
730 506 11 0.7942 0.01577 0.7639 0.8257
767 487 25 0.7534 0.01694 0.7210 0.7874
803 454 20 0.7203 0.01774 0.6863 0.7559
840 428 12 0.7001 0.01818 0.6653 0.7366
876 406 24 0.6587 0.01897 0.6225 0.6969
913 366 20 0.6227 0.01956 0.5855 0.6622
949 339 19 0.5878 0.02004 0.5498 0.6284
986 313 22 0.5465 0.02047 0.5078 0.5881
1022 281 23 0.5017 0.02081 0.4626 0.5442
1059 247 18 0.4652 0.02101 0.4258 0.5082
1095 224 25 0.4133 0.02107 0.3740 0.4567
1132 189 16 0.3783 0.02103 0.3392 0.4218
1168 163 15 0.3435 0.02092 0.3048 0.3870
1205 138 14 0.3086 0.02077 0.2705 0.3521
1241 114 13 0.2734 0.02057 0.2359 0.3169
1278 96 10 0.2449 0.02030 0.2082 0.2882
1314 84 8 0.2216 0.01997 0.1857 0.2644
1351 69 6 0.2023 0.01973 0.1672 0.2449
1387 60 6 0.1821 0.01941 0.1478 0.2244
1424 51 10 0.1464 0.01860 0.1141 0.1878
1460 32 10 0.1007 0.01753 0.0715 0.1416
1497 21 4 0.0815 0.01661 0.0546 0.1215
1533 15 1 0.0760 0.01637 0.0499 0.1159
1570 14 1 0.0706 0.01607 0.0452 0.1103
1606 11 2 0.0578 0.01550 0.0341 0.0978
1643 7 1 0.0495 0.01533 0.0270 0.0908
1679 6 1 0.0413 0.01483 0.0204 0.0835
1935 1 1 0.0000 NaN NA NA
Kaplan-Meier Curve:
Comment:
The estimated survival probabilities obtained using the KM formula are shown
here for [Link]. The Kaplan-Meier curve represents the survival probabilities
of the employees over time where the horizontal axis represents time and the
vertical axis represents the estimated probability. We show that the curve is starts
with a survival probability of 1 and then steps down to the other survival
probabilities as we move from one ordered turnover time to another that represent
a stepped downward curve which indicates a decline in survival probability (a
lower event rate).