27.03.2013 Views

SPSS® 12.0 Command Syntax Reference

SPSS® 12.0 Command Syntax Reference

SPSS® 12.0 Command Syntax Reference

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

1268 PROBIT<br />

• Keywords BY and WITH can appear in either order. However, both must follow the<br />

response-and-observation-count variables.<br />

Example<br />

PROBIT R OF N WITH X.<br />

• The number of observations having the measured response appears in variable R, and the<br />

total number of observations is in N. The predictor is X.<br />

Example<br />

PROBIT R OF N BY ROOT(1,2) WITH X.<br />

PROBIT R OF N WITH X BY ROOT(1,2).<br />

• Because keywords BY and WITH can be used in either order, these two commands are<br />

equivalent. Each command specifies X as a continuous predictor and ROOT as a categorical<br />

grouping variable.<br />

• Groups are identified by the levels of variable ROOT, which may be 1 or 2.<br />

• For each combination of predictor and grouping variables, the variable R contains the<br />

number of observations with the response of interest, and N contains the total number of<br />

observations.<br />

MODEL Subcommand<br />

MODEL specifies the form of the dichotomous-response model. Response models can be<br />

thought of as transformations (T) of response rates, which are proportions or probabilities<br />

(p). Note the difference in the transformations between the current version and the previous<br />

versions.<br />

• A probit is the inverse of the cumulative standard normal distribution function. Thus, for<br />

any proportion, the probit transformation returns the value below which that proportion<br />

of standard normal deviates is found. For the probit response model, the program uses<br />

Tp ( ) = PROBIT( p)<br />

. Hence:<br />

T( 0.025)<br />

= PROBIT( 0.025)<br />

= – 1.96<br />

T( 0.400)<br />

= PROBIT( 0.400)<br />

= – 0.25<br />

T( 0.500)<br />

= PROBIT( 0.500)<br />

= 0.00<br />

T( 0.950)<br />

= PROBIT( 0.950)<br />

= 1.64<br />

• A logit is simply the natural log of the odds ratio, p ⁄ ( 1 – p)<br />

. In the Probit procedure, the<br />

response function is given as Tp ( ) = log e(<br />

p ⁄ ( 1 – p)<br />

) . Hence:<br />

T( 0.025)<br />

= LOGIT( 0.025)<br />

= – 3.66<br />

T( 0.400)<br />

= LOGIT( 0.400)<br />

= – 0.40<br />

T( 0.500)<br />

= LOGIT( 0.500)<br />

= 0.00<br />

T( 0.950)<br />

= LOGIT( 0.950)<br />

=<br />

2.94

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!