There is absolutely no direct procedure option for calculating the KR21, which happens to be dependant on the (commonly unrealistic) assumption that every one products in the scale have equal trouble stages (i.e. equivalent proportions of 'successes' around the objects). The KR21 might be calculated in SPSS by using the https://kr2005938.full-design.com/kuder-richardson-reliability-an-overview-69479162