Statistics
Determining the Kappa coefficient for multiple graders can be a laborious task prone to calculation errors. However, the Kappa value is one of the best statistical methods for assessing agreement among telemedicine retinal image graders or fundus photography readers. An alternative is to look at percent agreement, which is less rigorous than Kappa. Below is a link to an online calculator to help you to compute Kappa (Fleiss' coefficient). It requires the total number of agreements across all readers for each category, but it saves you the effort of calculating the actual coefficient. Kappa Calculator Link
********************************************************************************************************
If you are looking for a tool to generate a Receiver Operating Characteristic (ROC) Curve, you can find one at the stAR website. The ROC curve is a graphical plot of the true positive rate (i.e., the sensitivity of a test) versus the false positive rate (one minus the specificity of a test). Ideally the test is run for paired data.
********************************************************************************************************
A random number generator can be useful in clinical research. One use is to assign experimental test subjects to one of two groups (depending on if the number is positive or negative or if the number is even or odd). A second reason is to issue randomly selected identity numbers to a test subject . This is for the protection of their privacy and to assure that study records remain HIPAA compliant.
A fully functioning random number generator widget from the Random.org website is presented below. The range of allowable numbers can be specified by entering values into the minimum and maximum fields. If additional complexity is required, the random generator can be used multiple times in series. For example, if the generator is used three times, a sequence of numbers such as 235-8-8923 or 9032-891-672 can be generated. This additional randomness can be useful for assigning ID numbers with less concern for accidentally generating identical numbers.