responsiveness

Responsiveness evaluates whether a questionnaire reflects real changes over time in the construct it measures. There are several ways to measure responsiveness. 

Standardised effect size (ES) is a measure of the magnitude of change. An example is Cohen’s d. Cohen’s d is calculated by dividing the difference between the means of the two groups by the pooled standard deviation. This provides a unitless measure of the effect size, which has the advantage of allowing comparisons between studies using different scales. Cohen’s d is interpreted as 0.20 = small; 0.50 = moderate; 0.80 = large.

Standardised Response Mean (SRM) is similar to effect size, but uses the standard deviation of the change scores and is calculated by dividing the mean change (average difference e.g., baseline and follow-up) by the standard deviation of the change (variability). The SRM is useful when you want to know how much the average change is relative to how much change varies among individuals.

ROC AUC: A ROC curve is a graph. The ROC curve is useful for identifying how well the data can accurately classify cases (e.g., high anxiety/low anxiety) and shows you the trade-off between a true positive and a false positive. The x-axis shows the False Positive Rate (how many mistakes the model makes in labelling negatives as positive: 1-specificity) and the y-axis shows the True Positive Rate (how good the model is at detecting the positives: sensitivity). The area under the curve (AUC) is translated to a number between 0 and 1 (1.0 means perfect classification, 0.5 means no better than random guessing). The larger the AUC, the better the classification.