TABLE 4.
Test Results = Positive (MLD) | Test Results = Negative (Not MLD) | ||
---|---|---|---|
Part 1: Definitions of Accuracy Indices | |||
MLD | True positive | False negative | Sensitivity: P(test+|MLD+) |
Not MLD | False positive | True negative | Specificity: P(test−|MLD−) |
Positive predictive value P(MLD+| test+)* | Negative predictive value P(MLD−| test−) | ||
MLD Status at Grade 3 | Test Positive (MLD) | Test Negative (Not MLD) | |
Part 2: Accuracy Based on Kindergarten Predictor Variables | |||
MLD | 26 | 15 | Sensitivity: 0.63 |
Not MLD | 61 | 104 | Specificity: = 0.63 |
Positive predictive value = 0.30 | Negative predictive value = 0.87 | ||
MLD Status at Grade 3 | Test Positive (MLD) | Test Negative (Not MLD) | Based on First Grade Data |
Part 3: Accuracy Based on First-Grade Predictor Variables | |||
MLD | 12 | 29 | Sensitivity: 0.29 |
Not MLD | 11 | 159 | Specificity: 0.94 |
Positive predictive value = 0.52 | Negative predictive value = 0.85 |
P(MLD+|test+) refers to the probability that a child is truly positive for MLD given that the test result for MLD is positive.