Psychological tests are tools used to measure differences between individuals or between the same individual under different circumstances. A psychological test is an objective and standardized measure of a sample of behavior. Before a test is released, researchers have to check its reliability and validity. By the term reliability it is meant consistency of scores obtained by the same person on different occasions. Validity of a test means that it actually measures what is meant to measure. Validity of a test tells the researcher what the test is measuring.
In this study, Rosenberg’s Self esteem scale is examined thoroughly by first looking at what self esteem is and how it is measured. The conceptualization of self esteem as a phenomenological construct seems to be linked with its measurement. There is plethora of research on the measurement of self esteem and some studies have shown that Rosenberg’s scale is valid and reliable for some samples but others find that because self esteem is an unstable characteristic its measurement cannot be reliable. Difficulties in defining self esteem are also linked to its methodological issues as some researchers define self esteem as evaluative and others as descriptive. Problems of measuring a construct like self esteem are identified in this study as well as research on what population Rosenberg’s self esteem scale is found to be valid and reliable is encountered.
Self esteem promotes behaviors, goals and gives individuals a sense of worth. Some theories suggest that self esteem is a basic human need or motivation. Maslows’ theory of hierarchy of needs included self esteem by describing two different types of esteem; the need for respect from others and the need for self respect. Without fulfillment of self esteem individuals, according to Maslow could not obtain self actualization. Robson (1989) defined self esteem as the sense of contentment and acceptance that results from a person’s appraisal of one’s own worth, attractiveness, competence and ability to satisfy ones aspirations. (Columbus, 2006) Rosenberg and social learning theorists defined self esteem in terms of a stable sense of personal worth.
These definitions have been applied to global and specific self esteem, meaning that global is an individual’s overall evaluation of himself and specific on a facet of the self such as academic competence. This is one of the problems of self report measurements. It looks at what the individual perceives of himself. If an individual perceives himself as competent then the test will result that the person has high self esteem regardless of how competent he really is. Rosenberg’s self esteem scale is an attempt to achieve a measure of global self esteem. There is research evidence that this scale is useful for getting a global sense of how a person feels about himself and is a good predictor of other measures of mental health such as depression and anxiety. Unfortunately research shows that it is not such a good predictor for specific domains like academic ability.
Self report is one of the most used methods for gathering information on children and adults and it’s consistent with theoretical concepts of how to obtain information about oneself. Of course, it’s not without its drawbacks. There are issues on how accessible this information is in the memory, the situations or contexts a person is when they are answering and other factors that might influence how one answers questions about himself.
As mentioned earlier cognitive and affective processes are involved in how information is processed through the memory and how it’s retrieved at a particular moment. A method for finding reliability is based on the consistency of responses to all items in the test. Due to the cognitive and affective processes, there are biases such as self-deception that occur when accessing information. The context and the emotional state where someone is asked to answer self report questionnaire on self esteem involves the situational factors that might influence responds about the self. If for example, someone is asked to answer about his self esteem after getting a promotion it will elicit his response to feel better with himself thus with more self esteem. Self presentation also impedes a person’s belief about himself as he might want to be perceived by the researchers as someone with self esteem. It is important to know if individuals are responding honestly as social desirability is a factor that alters feelings. These factors may be biasing the responses they are giving.
Rosenberg’s self esteem scale is a ten item scale that asks individuals to rate themselves using four categories (strongly agree to strongly disagree) on statements regarding their sense of self worth or confidence. The scale is perhaps the most famous and widely used measure of self esteem. It is used with children and adults and is considered to be reliable and valid measure of self esteem. The SES has been validated for use with substance abusers and other clinical groups, and is regularly used in treatment outcome studies. The scale has been validated for use with both male and female adolescent, adult and elderly populations. It has been used with francophone populations but studies about other cultural groups are debateful. That might be the case because of the fact that the concept of self is highly dependent on verbal behavior and social norms in each society. The complexity of the concept as well as the lack of theoretical agreement and empirical data has resulted in accepting a general definition of self esteem, which is mirrored in the broad use of Rosenberg’s’ self esteem scale. (Columbus, 2006)
Although Rosenberg’s self esteem scale has excellent psychometric properties it constitutes a unidimensional scale that does not capture the complexity of self esteem. Unidimensional scales are useful for gathering information on general mental health of individuals. There are a lot of scales that measure aspects of self esteem. Researchers should use measures that are specific to the issue they are investigating whether it is a global sense of self worth or specific evaluation of competencies across a wide spectrum of behaviors. There are allot of studies examining Rosenberg’s self esteem scale and their results vary. An important condition influencing the size of a reliability coefficient is the nature of the group on which reliability is measured.
A study that examined further support for multidimensionality within the Rosenberg self-esteem scale suggested that Tafarodi and Swann (1995) had another perspective on the relationship of Rosenberg’s self esteem scale and the concept of global esteem. They thought that global self esteem is comprised by two separate and different concepts; self competence and self liking.
A study examining the psychometric properties of the Rosenberg Self esteem scale in Chinese acute coronary syndrome patients (2006) suggest that the RSES has been found associated with depression. The RSES has also been widely used to gain insight into a diverse range of areas of substantive psychological interest including aspects of gender (Huyck, 1991), aggression and family dynamics (Haj-Yahia, 2001), perfectionism (Stumpf & Parker, 2000) and neuroticism (Pullman & Allik, 2000). Self esteem is found to be related to basic dimensions of personality as individuals with high self esteem tend to score high on extraversion, agreeableness and openness to experience. A more complex issue is the relation of self esteem and narcissism, as they have been found to correlate. The RSES has been found to correlate well with tests of self concept (Beck, Steer, Epstein, & Brown, 1990) thus supporting the construct validity of the instrument. Rosenberg’s self esteem scale has been translated into a lot of languages and is used in diverse populations.
One such example is the translation and validation of the Rosenberg self esteem scale in Spanish. The study consisted of the translation, adaption and analysis of the psychometric properties of the scale in a sample of university students. The results showed that the scale presents a unifactorial structure with good levels of internal consistency and stability. Test- retest correlation was good, supporting the reliability of the scale.
A recent study (2009) examining the Psychometric Properties of the Rosenberg Self-Esteem Scale in African American Single Mothers shows that the scale can be released to other cultural groups. The purpose of this study was because self-esteem has been identified as an important factor in the mental health of African American women. African American women unfortunately experience multiple oppressions related to both racism and sexism, which has an impact on the formation and maintenance of self-esteem. Women with this kind of background, and especially single mothers with low income are correlated with poor self-esteem (Jackson, 2003) The RSE Scale showed adequate internal consistency with an alpha coefficient of .83. Two factors that accounted for a total of 54.7% of the variance were extracted. Self-esteem showed a strong negative relationship with both depressive symptoms and negative thinking. This study provides support for the internal consistency of the RSE Scale and partial support for its construct validity in this population. The RSE appears to represent a bidimensional construct of self-esteem for African American women, with the cultural influences of racial esteem and the rejection of negative stereotypes forming a separate and distinct aspect of this concept.
As seen above, there is empirical evidence that shows that socioeconomic status plays role in self esteem scores. It is very weakly correlated to self esteem but it seems that wealthy individuals have a bit more self esteem than working class individuals. Researchers who study self esteem generally assume that it’s a stable trait that predicts future behavior. Research indicates that although self esteem levels can be temporarily elevated or depressed in everyday life studies show the stability of self esteem across years. Test retest is a method for finding the reliability of test scores by repeating the identical test on a second occasion. Test retest correlation of self esteem is comparable to those found for more basic dimensions of personality such as extraversion or neuroticism. There has been research on the stability of self esteem and it seems that it is high in childhood, falls in adolescence and increases during adulthood.
A study made in 2006 by Quilty & Oakman& Risko suggest that although Rosenberg’s self esteem scale was developed as a global self esteem factor, there is mixed support for its unifactorial structure. Many investigation of the structure of the scale used factor analysis. Factor analysis was developed as a mean of identifying psychological traits and is used for analyzing the interrelationships of behavior data. Hensley and Roberts (1976) found that an exploratory factor analysis of the scale results in two factors, consisting of the positively and negatively worded items which reflected an underlying response set. A study looking at the norms and construct validity of RSES in Canada suggests that Kaplan and Pokomey (1976) found that while there was a strong unitary factor in the scale, research showed a two factor solution with positively worded items and negatively worded items. These factors were called defense of individual self worth and self derogation accordingly. In 1990 another study looked into the two factor structure of Rosenbergs self esteem scale and studied 1,700 adults in the USA and found the same results. They also argued that the scale can be regarded psychometrically coherent with a unitary structure. These findings indicate that the RSES is a reliable and potentially valid scale for use with Canadian high school students of all ages (Bagley & Bolitho & Bertrand ,1997)
Another study compared Rosenberg’s self esteem scale with Coppersmiths’ self esteem inventory with people with eating disorders. Researchers wanted to see which test would fit best in measuring self esteem so they measured the validity of both tests. The study found that Rosenberg’s self esteem scale had better construct and convergence validity than the Coppersmith self esteem inventory at the particular sample.
It is obvious that reliability and validity in some samples might not be granted. Reliability is the consistency of scored obtained by the same sample when they are reexamined with the same test on different occasions. Test reliability indicates the extent to which individual differences in test scores are attributable to ‘true’ differences in the characteristics under consideration and the extent to which they are attributable to chance errors. When the differences are not ‘true’ they are called error variance and this is a possibility when measuring self esteem. The researcher would have to test the correlation coefficient of two sets of scores.
As we can see from studies mentioned when researching the self, the methodological issues are difficult. The researchers should examine each tests reliability and validity before releasing it to the public. The problems of measuring a construct like self-esteem apply on nearly all of the studies that measure elements of the personality. Some researchers assume that phenomena that occur at the cognitive level will be manifested by behaviors that can be observed and measured. Other researchers suggest that behaviors should not be used to measure personality traits because we are unable to determine what cognitive processes might be affecting that trait. They argue that constructs like self-esteem should not be studied because there is no clear way to measure them or even know if you are measuring the right thing.
As the majority of the theories suggest, individuals develop their self and self-concept through social interaction. There may be some innate cognitive process that is modified by social interaction or just social interaction, but either way people display behaviors and express opinions about who they are, from a very young age. For this reason it is important that psychologists be able to study and measure this phenomenon. Future work needs to address the issues of theory, definition, and methodology that have remained unresolved across the centuries. A consensus on these issues will lead to better measurement and a better understanding of how self-esteem influences the mental health and behavior of individuals.