![]() ![]() where each observed predicted probability is used as a cutoff value for classification). That resolution shows you how to see the sensitivity and false positive rates (1 - specificity) for all observed cutoff values (i.e. If you do not have a specific cutoff value in mind, you may find Technote #1479847 ("C Statistic and SPSS Logistic Regression") to be helpful. This process of testing sensitivity for another input (say cash flows growth rate) while keeping the rest of inputs constant is repeated until the sensitivity figure for each of the inputs is obtained. You can choose a different cutoff value for the classification by entering a value in the "Classification cutoff" box in the lower right corner of the Options dialog of Logistic Regression. The sensitivity is calculated by dividing the percentage change in output by the percentage change in input. ![]() So, the percentage of correct classification figures represent the specificity and sensitivity when the cutoff value for the predicted probability =. Otherwise, the case is classified as the non-target event. ![]() By default, a case is classified as the target category if the probability of the target event is greater than or equal to. The percentage correct for the second category, the target category, is the sensitivity (which is also usually expressed as a proportion). The percentage correct for the first category is the specificity, although this is usually expressed as a proportion. There will be a "Percentage Correct" column with the percentage of correct classifications for each of the DV categories. In binary logistic regression, the higher value of the DV is necessarily the category whose probability is predicted by the model (i.e., the target category) and will be the second row and column of the classification table. In the classification table in LOGISTIC REGRESSION output, the observed values of the dependent variable (DV) are represented in the rows of the table and predicted values are represented by the columns. Sensitivity and Specificity are displayed in the LOGISTIC REGRESSION Classification Table, although those labels are not used. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |