Reliability and Validity of Survey Scales

In this paper, we answered two questions: What is the reliability of a response scale in a question? What is the validity of a response scale in a question? The purpose of this paper is to present practical tools for measuring the reliability and validity of response scales used in written survey. Reliability measures consistency and validity measures precision. Our objective is to determine the reliability and validity of Likert and non-Likert scales used in research instrument. The data came from the numerical values of each type of scale. The Likert-type of scales include (1,2,3,4,5), (1,2,3,4,5,6,7) and (1,2,3,4,5,6,7,8,9,10). Non-Likert scale was (0,1,2,3). Reliability was measured by the estimated of under system analysis. The response space was proxied as a system to create a range between maximum and minimum values in the scale. Validity was tested by using the Fisher transformation of the estimated Z score of series. Empirical evidence shows that non-Likert scale (0,1,2,3) is 92% reliable while the Likert-type of scale had 90, 89, and 88% reliability. Validity test showed that non-Likert scale was 93% reliable, while the Likert-type scale had 89, 61, and 57% precision. Through Monte Carlo simulation and NK landscape method for optimization, the ability of information retention for non-Likert scale was 0.96 and 0.73, 0.75, and 0.77 for Likert scales. We standardize the scale efficacy in a 5.0 system, the non-Likert scale is 4.73 and 2.35, 2.45, and 2.41 for Likert scales.In this paper, we answered two questions: What is the reliability of a response scale in a question? What is the validity of a response scale in a question? The purpose of this paper is to present practical tools for measuring the reliability and validity of response scales used in written survey. Reliability measures consistency and validity measures precision. Our objective is to determine the reliability and validity of Likert and non-Likert scales used in research instrument. The data came from the numerical values of each type of scale. The Likert-type of scales include (1,2,3,4,5), (1,2,3,4,5,6,7) and (1,2,3,4,5,6,7,8,9,10). Non-Likert scale was (0,1,2,3). Reliability was measured by the estimated of under system analysis. The response space was proxied as a system to create a range between maximum and minimum values in the scale. Validity was tested by using the Fisher transformation of the estimated Z score of series. Empirical evidence shows that non-Likert scale (0,1,2,3) is 92% reliable while the Likert-type of scale had 90, 89, and 88% reliability. Validity test showed that non-Likert scale was 93% reliable, while the Likert-type scale had 89, 61, and 57% precision. Through Monte Carlo simulation and NK landscape method for optimization, the ability of information retention for non-Likert scale was 0.96 and 0.73, 0.75, and 0.77 for Likert scales. We standardize the scale efficacy in a 5.0 system, the non-Likert scale is 4.73 and 2.35, 2.45, and 2.41 for Likert scales.JEL Code: C12, C13, C15, C18, C83, C9