Application of decision theory to DUI assessment

Area of Science:

• Decision science
• Forensic psychology
• Criminal justice

Background:

• Screening tools are crucial for identifying individuals driving under the influence (DUI).
• Evaluating the effectiveness of these screening instruments requires robust analytical methods.
• Traditional methods may not fully capture the nuances of screening performance.

Purpose of the Study:

• To apply decision theory principles to the evaluation of DUI offender screening instruments.
• To investigate the validity of a structured interview and a survey instrument for DUI screening.
• To introduce and demonstrate the utility of the "cost ratio" as a novel measure for assessing examiner bias in DUI screening.

Main Methods:

• Utilized graphical analysis with Relative Operating Characteristic (ROC) curves.
• Examined the relationship between true-positive and false-positive rates across different decision cut-off scores.
• Quantitatively described instrument performance using test sensitivity and examiner bias parameters.

Main Results:

• ROC curves provide a visual representation of screening tool performance.
• Test sensitivity and examiner bias are key parameters for evaluating screening instruments.
• The "cost ratio" offers an improved method for quantifying examiner bias in DUI screening.

Conclusions:

• Decision theory provides a valuable framework for evaluating DUI screening programs.
• ROC analysis and the "cost ratio" enhance the quantitative assessment of screening instruments.
• The findings support more accurate and objective evaluation of tools used for DUI offender screening.