Logistic Regression (Multiple) LOGIST2

Exercise:
 The purpose of the study is to investigate the association between colon cancer prognosis, and surgical and histopathological findings. Perform logistic regression analysis. The objective variables (dependent variable) are defined as colon cancer prognosis, ‘survival = 1, death = 0’. Explanatory variables (independent variable) are defined as ‘cancer invasion depth (1 to 5 grades), serous membrane invasion = 0 or 1, and hepatic metastasis = 0 or 1

Comparative Verification Test

Likelihood ratio test, Wald test, and 95% confidence intervals for odds ratio: StatMate and SPSS, perfect match.
Akaike’s information criterion (AIC): StatMate and EZR, perfect match.
Note: Regarding StatMate’s global null hypothesis and SPSS’s omnibus test
StatMate performs the stepwise method by aligning various test methods (-2LOG L likelihood ratio test, score test, Wald test) + 95% confidence intervals + Akaike’s information criterion (AIC) + Schwartz C value. The omnibus likelihood ratio test in SPSS is said to be the most reliable of these.

Reports of SPSS & EZR

Since the item names in StatFinale are different from those in SPSS, we have created tables in SPSS report format to make it easier to add and remove variables. “Wald” in SPSS is the Wald test statistic (χ2 value), and the P value is calculated using CHIDIST(Wald value, degrees of freedom). “Exp(B)” in SPSS is the odds ratio (the ratio of the probability that an event occurs to the probability that it does not occur).
The EZR values ​​are slightly different from StatFinale and SPSS, but are almost the same. EZR displays the AIC value, which is a perfect match with StatFinale.

※Click pics to enlarge.

pagetop