Stat Computing > SPSS > Textbook Examples > Applied Logistic Regression, 2nd, by Hosmer and Lemeshow

SPSS Textbook Examples
Applied Logistic Regression, Second Edition, by Hosmer and Lemeshow
Chapter 3: Interpretation of the fitted logistic regression model

page 51 Table 3.2 Cross-classification of age dichotomized at 55 years and chd for 100 subjects

get file='chdage.sav'.

recode age (55 thru highest=1) (else=0) into aged.
execute.

CROSSTABS
  /TABLES=chd  BY aged
  /FORMAT= AVALUE TABLES
  /CELLS= COUNT.

page 52 Table 3.3 Results of fitting the logistic regression model to the data in Table 3.2.

LOGISTIC REGRESSION VAR=chd
  /METHOD=ENTER aged.
 

page 56 Table 3.5 Cross-classification of hypothetical data on race and chd status for 100 subjects.

data list list / race chd cnt.
begin data.
 1 1 5
  2 1 20
  3 1 15
  4 1 10
  1 0 20
  2 0 10
  3 0 10
  4 0 10
end data.
execute .

weight by cnt.

CROSSTABS
  /TABLES=chd BY race
  /FORMAT= AVALUE TABLES
  /CELLS= COUNT .

LOGISTIC REGRESSION VAR=chd
  /METHOD=ENTER race
  /CONTRAST (race)=Indicator(1)
  /PRINT=SUMMARY CI(95).

NOTE: The above code also gives the coding scheme shown in Table 3.6.

page 58 Table 3.7 Results of fitting the logistic regression model to the data in Table 3.5 using the design variables in Table 3.6.

NOTE: The above code also gives the output shown in Table 3.7.

page 59 Table 3.8 Specification of the design variables for race using deviation from means coding.

LOGISTIC REGRESSION VAR=chd
  /METHOD=ENTER race
  /CONTRAST (race)=Deviation(1)
  /PRINT=SUMMARY CI(95).

page 60 Table 3.9 Results of fitting the logistic regression model to the data in Table 3.5 using the design variables in Table 3.8.

NOTE: The above code also gives the output shown in Table 3.9.

NOTE: To get the values listed in the column labeled z, you need to take the square root of the Wald statistics given in the SPSS output.

page 67 Table 3.10 Descriptive statistics for two groups of 50 men on age and whether they had seen a physician (PHY) (1 = yes, 0 = no) within the last six months.

NOTE: These data are hypothetical and are not available.

page 69 Table 3.11 Results of fitting the logistic regression model to the data summarized in Table 3.10.

NOTE: These data are hypothetical and are not available.

page 72 Table 3.12 Estimated logistic regression coefficients, deviance, and the likelihood ratio test statistic (G) for an example showing evidence of confounding but no interaction (n = 400).

NOTE: These data are hypothetical and are not available.

page 73 Table 3.13 Estimated logistic regression coefficients, deviance, and the likelihood ratio test statistic (G) for an example showing evidence of confounding and interaction (n = 400).

NOTE: These data are hypothetical and are not available.

page 77 Table 3.14 Estimated logistic regression coefficients, deviance, and the likelihood ratio test statistic (G), and the p-value for the change for models containing lwd and age from the low birth weight data (n = 189).

NOTE: We have run the logistic regression models from the largest to the smallest so that the difference between the larger and the smaller model can be determined. This is the reverse of the presentation in the table in the book.

NOTE: To get the ln[l(beta)], divide the -2 log likelihood given in the output by -2. To obtain the values of G, subtract the value of ln[l(beta)]. from that of the model with one more term in it (for example,  -117.34-(-113.12)=8.44).

Get file='lowbwt.sav'.

compute lwd=(lwt<110).
compute lwdage=lwd*age.
execute.

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER lwd age lwdage.
 

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER lwd age.

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER lwd.
 

NOTE: To get the model with only the intercept, you need to create a variable equal to one and use that as the dependent variable.

compute x = 1.

 LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER x
  /ORIGIN.
 

page 78 Figure 3.3 Plot of the estimated logit for women with LWD = 1 and for women with LWD = 0 from Model 3 in Table 3.17.

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER lwd age lwdage
  /SAVE PRED.
 

GRAPH
  /SCATTERPLOT(BIVAR)=age WITH pre_1.

page 78 Table 3.15 Estimated covariance matrix for the estimated parameters in Model 3 of Table 3.14.

NOTE: There are likely typos in this table.

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER lwd age lwdage
  /PRINT=corr.
 

constant/constant: (.910)**2 = .8281 
constant/lwd: (.910)*(1.725)*(-.528) = -.828828 
constant/age: (.910)*(.040)*(-.978) = -.0355992 
constant/lwd*age: (.910)*(.076)*(.512) = .03603236 
lwd/lwd: (1.725)**2 = 2.975625 
lwd/age: (1.725)*(.040)*(.516) = .035604 
lwd/lwd*age: (1.725)*(.076)*(-.977) = -.1280847 
age/age: (.040)**2 = .0016 
age/lwd*age: (.040)*(.076)*(-.524) = -.001593 
lwd*age/lwd*age: (.076)**2 = .005776

page 79 Table 3.16 Estimated odds ratios and 95% confidence intervals for lwd, controlling for age.

NOTE: We were unable to reproduce this table.

page 80 Table 3.17 Cross-classification of low birth weight by smoking status.

CROSSTABS
  /TABLES=low BY smoke
  /FORMAT= AVALUE TABLES
  /CELLS= COUNT.
 

page 81 Table 3.18 Cross-classification of low birth weight by smoking status stratified by race.

CROSSTABS
  /TABLES=low BY smoke BY race
  /FORMAT= AVALUE TABLES
  /CELLS= COUNT.
 

page 82 Table 3.19 Tabulation of the estimated odds ratios, ln(estimated odds ratios), estimated variance of the ln(estimated odds ratios), and the inverse of the estimated variance, w, for smoking status within each stratum of race.

NOTE: The estimated variance of the ln(estimated odds ratios), and the inverse of the estimated variance, w, were not calculated  because they were needed only to do a hand-computation.

compute race1=0.

recode race1 (0=1) (1,2=2) (3 thru 15=3) (16 thru highest=4).

recode race (2=1) (else=0) into race2.
recode race (3=1) (else=0) into race3.
compute race1sm=race1*smoke.
compute race2sm=race2*smoke.
compute race3sm=race3*smoke.
execute.

NOTE: Values for White:

LOGISTIC REGRESSION VAR=low
  /METHOD=ENTER race2 race3 race2sm race3sm smoke.