ABSTRACT
The key difference between Lambda and Cramer’s V is there is an antecedent(predictor) and consequent(predicted) variable based upon a chronological relationship between the variables—one variable comes before the other variable in time. The value of Lambda can range from 0 to 1. In essence the statistic is a measure of the relative reduction in prediction error of one variable when social worker knows the value of the other variable. If the prediction of job type had been based on chance alone, the chance of making the correct prediction would have been 25%, resulting in 75% errors. The Lambda coefficient indicates that knowing the type of university the student graduated from will reduce the error of prediction in the type of social work job where they are employed by 26%. A scatterplot of the predicted standardized value against the standardized residual value helps to visualize homoscedasticity between predicted dependent variable scores and errors of that prediction.
