The Cronbach's coefficient (Cronbach's alpha) is a statistic that refers to the mean of the folded-half reliability coefficients obtained from all possible ways of dividing items in a scale, and it is the most commonly used measure of reliability. It was first named by American educator Lee Cronbach in 1951. Where K is the number of samples, σ2X is the variance of the total sample, and σ2Yi is the variance of the currently observed sample.

Cronbach's coefficient

Usually the Cronbach alpha coefficient has a value between 0 and 1. If the X coefficient does not exceed 0.6, the internal consistency reliability is generally considered insufficient; when it reaches 0.7-0.8, it indicates that the scale has considerable reliability, and when it reaches 0.8-0.9, it indicates that the reliability of the scale is very good.

An important characteristic of Cronbach's alpha coefficients is that their value increases with the number of items on the scale, so Cronbach's alpha coefficients may be artificially and inappropriately inflated by the inclusion of redundant measurement items in the scale.

There is another type of coefficient that can be used alongside the Cronbach alpha coefficient. The coefficient can help evaluate whether the calculation of the mean masks certain irrelevant measurement items in the calculation of the Cronbach alpha coefficient.

Different researchers have different views on the cut-off value of the reliability coefficient, and some scholars believe that the Cronbach alpha coefficient should be at least 0.8 to be acceptable for basic research, at least 0.7 to be acceptable for exploratory research, and only 0.6 to be acceptable for practical research.

Definition

Standardized Cronbach alpha coefficient definition, if a scale has n questions, and the average correlation coefficient between the questions is r, the standardized alpha coefficient for this scale is, α=nr/[(n-1)r+1].