The RSES is a widely used self-esteem measurement instrument based on self-reporting by subjects. The tool’s success is attributed to its brevity and a format that is simple and user-friendly. “RSES consists of ten items which the test taker answers on a four-point scale ranging from strongly agree to strongly disagree” (Alessandri et al., 2015). An assessment of the validity of the RSES would involve ascertaining the degree to which the instrument measures an individual’s self-esteem. If one were to create a new self-esteem measurement instrument, it would need exhibit both reliability and validity. Using the RSES as an example, this paper will illustrate how one would determine an instrument’s validity.
Principal Component Analysis
An instrument measuring self-esteem will typically have a number of items that ascertain various aspects of an individual’s self-worth. The RSES, for example, has ten items. Reducing this large set of variables is necessary if one is to determine the instrument’s validity. Principal component analysis reduces the large set to a smaller one that still accounts for most information on the larger variant (Bro and Smilde, 2014). In the case of the RSES, the wording of the items is the appropriate parameter for delineation. The components that would be analyzed in the case of RSES are the positive and negative wording of items. According to a study by Tinakon and Nahathai (2012), items 1, 2, 4, 7, and 10 in the RSES are positively worded while items 3, 5, 6, and 9 are negatively worded. This analysis makes it easier to explore and visualize data. An intercorrelation analysis after this step would help one determine how the various items of the tool differentiate between each other. The distinction between items indicates proper design since test takers would appreciate their relevance and clarity (Kline, 2005).
Empirical Validity
The appropriate methodology to use when determining an instrument’s validity would be to compare it with different measure and determine convergent validity using the Pearson product coefficient of correlation. In the context of self-esteem, measures such as depression, anxiety, and stress subscales would form a good basis for empirical determination. “Empirical validity is demonstrated when an instrument correlates highly with other tools which it shares overlapping items or constructs” (Rizwan et al., 2017).
References
Alessandri, G., Vecchione, M., Eisenberg, N., " Łaguna, M. (2015). On the factor structure of the Rosenberg (1965) General Self-Esteem Scale. Psychological Assessment, 27(2), 621.
Bro, R., " Smilde, A. K. (2014). Principal component analysis. Analytical Methods, 6(9), 2812-2831.
Kline, T.J.B. (2005). Psychological testing: a practical approach to design and evaluation. Thousand Oaks, CA: Sage Publications.
Rizwan, M., Malik, M. S., Malik, N. J., " Siddiqui, R. S. (2017). Urdu Rosenberg self–esteem scale: an analysis of reliability and validity in Pakistan. Sociology International Journal, 1(2), 1–0. https://doi.org/10.15406/sij.2017.1.00010
Tinakon, W., " Nahathai, W. (2012). A Comparison of Reliability and Construct Validity between the Original and Revised Versions of the Rosenberg Self-Esteem Scale. Psychiatry Investigation, 9(1), 54–58. https://doi.org/10.4306/pi.2012.9.1.54
