Skip to main content

Table 4 The bias and RMSE in parentheses of the Rogan-Gladen adjusted prevalence estimates when targeting a specificity of 0.995. The method(s) with the smallest bias in each scenario or equivalent after rounding are bolded

From: Estimating cutoff values for diagnostic tests to achieve target specificity using extreme value theory

 

Scenario A

Scenario B

 

n=50

n=200

n=50

n=200

Prevalence=0.05

     Empirical

0.02

(0.0328)

0.01

(0.0212)

0.02

(0.0315)

0.00

(0.0119)

     Normal

0.04

(0.0406)

0.04

(0.0362)

0.03

(0.0300)

0.01

(0.0145)

     Log Normal

0.03

(0.0317)

0.03

(0.0291)

0.02

(0.0213)

0.02

(0.0159)

     MAD

0.09

(0.0911)

0.09

(0.0878)

0.14

(0.1374)

0.14

(0.1361)

     Log MAD

0.05

(0.0499)

0.05

(0.0461)

0.08

(0.0754)

0.07

(0.0695)

     Pareto 0.9

0.01

(0.0453)

0.00

(0.0204)

0.02

(0.0247)

0.00

(0.0074)

     Pareto 0.95

0.02

(0.0382)

0.00

(0.0213)

0.02

(0.0346)

0.00

(0.0096)

     Hybrid Empirical

0.02

(0.0322)

0.01

(0.0212)

0.02

(0.0311)

0.00

(0.0119)

     Hybrid Pareto 0.9

0.02

(0.0478)

0.00

(0.0205)

0.02

(0.0255)

0.00

(0.0074)

     Hybrid Pareto 0.95

0.02

(0.0395)

0.00

(0.0214)

0.03

(0.0348)

0.00

(0.0096)

Prevalence=0.30

     Empirical

0.02

(0.0563)

0.01

(0.0396)

0.01

(0.0309)

0.00

(0.0161)

     Normal

0.03

(0.0390)

0.03

(0.0295)

0.02

(0.0244)

0.01

(0.0125)

     Log Normal

0.03

(0.0418)

0.02

(0.0290)

0.02

(0.0173)

0.01

(0.0122)

     MAD

0.07

(0.0676)

0.06

(0.0646)

0.10

(0.1012)

0.10

(0.1003)

     Log MAD

0.04

(0.0420)

0.03

(0.0343)

0.06

(0.0553)

0.05

(0.0512)

     Pareto 0.9

0.01

(0.0976)

0.00

(0.0438)

0.01

(0.0293)

0.00

(0.0105)

     Pareto 0.95

0.02

(0.0749)

0.00

(0.0447)

0.01

(0.0349)

0.00

(0.0154)

     Hybrid Empirical

0.02

(0.0554)

0.01

(0.0396)

0.01

(0.0305)

0.00

(0.0161)

     Hybrid Pareto 0.9

0.01

(0.0922)

0.00

(0.0438)

0.01

(0.0298)

0.00

(0.0105)

     Hybrid Pareto 0.95

0.02

(0.0722)

0.00

(0.0447)

0.01

(0.0350)

0.00

(0.0154)

  1. Abbreviations: MAD mean absolute deviation