Kuder Richardson method, it is the same formula as Cronbach's Alpha except that the latter is expressed for continuous items and Kuder Richardson for dichotomous items.
There are several procedures for calculating the reliability of a measurement technique. All of these use formulas that produce reliability coefficients. These factors can range from 0 to 1. Where a coefficient of 0 means zero reliability and 1 represents a maximum of optimal reliability (total reliability).
The closer the coefficient is to zero (0), the greater the error in the measurement and the closer it is to 1 the better the measurement. For the application of this method to be effective, the following assumptions must be followed:
-The denominator is the variance of the total test scores.
-The numerator is the true variance, or the sum of covariances of the items.
-If the items do not discriminate, their standard deviations will be small, the numerator will be lower and therefore the reliability will also be lower.
If the standard deviations are large but the items are not related, the reliability will decrease, because this non-relation between the items means that the total scores are not differentiated.
The strength of the content is complex to obtain. First, it is necessary to review how the variable has been used by other researchers. And based on this review, a universe of possible items will be developed to measure the variable and its dimensions.
