Scribd
Upload
76 views39 pages

05 - Spps Logistic Regression

The document discusses logistic regression and odds ratios (OR). Logistic regression can be used when the outcome is binary and one or more predictors are continuous, binary, or categorical. It enables calculating the OR, which estimates the strength of the association between predictors and the outcome while controlling for other predictors. The multiple logistic regression analysis eliminates statistically insignificant predictors and provides ORs that are adjusted for confounding from other predictors included in the model.

Uploaded by

ESSA GHAZWANI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
76 views39 pages

05 - Spps Logistic Regression

The document discusses logistic regression and odds ratios (OR). Logistic regression can be used when the outcome is binary and one or more predictors are continuous, binary, or categorical. It enables calculating the OR, which estimates the strength of the association between predictors and the outcome while controlling for other predictors. The multiple logistic regression analysis eliminates statistically insignificant predictors and provides ORs that are adjusted for confounding from other predictors included in the model.

Uploaded by

ESSA GHAZWANI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 39

• Both the exposure variable and the outcome

variable are continuous. ( body weight- blood pressure)


• The exposure variable can be either continuous or
binary but the outcome variable is need to be binary only (dichotom).
(eg. Body weight- existing cardiovscular disease). It enables
statistical calculation of Odds (OR). In case of the multivariable logistic
regression using OR calculation, we can select the statistically significant
effects out of the multiple exposure factors that can have reliable
correlation with the outcome. OR is already cleared from the confounding
effect of every exposure factors taken in consideration in the analysis.
 What is the correlation in between the gratitude
of the debt (creddebt) to the defaults in payback
(default)
Dependent variable: Dependent
Independent variable: Covariates
B Correlational coefficient
Sig. p-value
Exp(B) OR
 What is the correlation between the amount of
the income and the default debt payout?
 What is the difference if we take in consideration
the gratitude of the debt and the amount of the
income at the same time?
 OR calculation of multiple logistic regression
calculates with the confounding effects of the
exposure variables.
 creddebt
 income
 default
 Database: 3rd_fact_demo_vegl.sav

Analysis of the correlation between smoking and


lung cancer development: do we get different
result from the OR calculation with the 2X2
contingency analysis?
Risk Estimate
95% Confidence Interval
Value Lower Upper
Odds Ratio for Smoking (Yes 8,750 6,257 12,236
/ No)
For cohort Lung cancer = 3,392 2,704 4,254
Yes
For cohort Lung cancer = No ,388 ,336 ,448
N of Valid Cases 800
SPSS program is able to take the interaction between the
variables in account, and to clear the results from the effect of
them.
SPSS program
eliminates the
statistically not
significant factor!
The programme eliminated the factor that
had no significant correlation with the
outcome
The program eliminates those
factor, which has the less effect on
the outcome!
 Endpoints eg:
 Recovery
 Disappearance of the symptoms
 Occurrence of an event
 Development of a disease
 Death

 Status: people reaching the defined endpoint

 Statistical analysis: Log rank


 How does the new operation method change the
survival chance of the patients?

Database: pain_medication.sav (Sample database of SPSS)


When it has reached the
end point it is 1, when
didn’t, it is 0.
 database: daganat_tuleles.sav

 How the survival of tumors are influenced by


„high risk” alleles?
 How the location of tumour influence the
survival?
 database: daganat_tuleles.sav

You might also like