On the validity of the state-like component of global self-esteem: Relationships with implicit self-esteem and work-related variables using different Latent State-Trait models

The present PhD dissertation includes three studies, which are organized into two chapters: The first, containing two studies, and the second containing one study. Overall, they aim to investigate the validity of the state-like variance of Global Self-Esteem (GSE). Along with this substantive focus, there is a methodological link connecting the studies: Indeed, in all studies, I used models belonging to the Latent State-Trait framework. Currently, GSE is one of the most studied variables in personality and social sciences research. Yet, the instability of GSE has also received much attention in recent years. Indeed, while several authors have reported a high percentage of trait variance in GSE, a large body of research has attested that GSE state variance is also significantly different from zero. This means that, as with most psychological constructs, GSE is composed by a stable and enduring “trait-like” component, as well as by an ephemeral and volatile “state-like” component. Given that these two GSE components are, by definition, orthogonal, they deserve to be studied with equal attention. However, most attention was given to the trait-like component of GSE; thus, in this dissertation we faced two debates surrounding the topic of the state-like variance of GSE. In the first chapter, State-Trait Decomposition of Name Letter Test Scores and Relationships with Global Self-Esteem, I carried out two longitudinal studies (participants in both studies were students) to investigate whether state variance in GSE is a measure of implicit self-esteem. This objective led me to investigate the validity of the Name Letter Test (NLT) as an implicit measure of state GSE. The NLT is an implicit measure assessing the likeability of the letters of one’s own first and family names. For several years, it has been assumed that one’s score in NLT should be a manifestation of state GSE. However, no studies empirically verified this assumption. Our aim was to use adequate statistical tools, belonging to the Latent State-Trait framework, to ascertain whether the NLT is a valid instrument for assessing the state-like component of GSE. In the second chapter, A Latent State-Trait Analysis of Global Self-Esteem in a Sample of Military Cadets: A Reconsideration of the State Component of Global Self-Esteem, I conducted one longitudinal study to attest the relevance of the GSE state-like component in organizational settings. Indeed, I noted that in organizational literature there is a great theoretical interest in state-like constructs, but surprisingly few studies used Latent State-Trait models, and therefore most of the evaluations about the trait- state-like components of the constructs were made simply by relying on theory. This procedure may be detrimental, because the literature demonstrated that almost every psychological construct may have both a significant trait variance and a significant state variance. Therefore, my aim was to show that, in organizational contexts, the state-like component of an enduring construct like GSE should be of interest for researchers and practitioners. In this study (again by using Latent State-Trait models), I verified whether the state component of GSE correlates with the state components of several work-related variables. This procedure allows the validity of the state-like component of GSE to be attested, and may therefore be a good starting point for reconsidering the importance of the state-like component of GSE in organizational studies. Overall, the rationale of the whole dissertation was to highlight some limitations in previous research addressing the validity of the state-like component of GSE, in particular (a) how and why the relationship between the state-like component of GSE and Name Letter Test scores has not been properly investigated, and (b) how and why the relevance of the state-like component of GSE in organizational psychology literature has been underestimated for some considerable time. Yet, the aim of the whole dissertation was first of all to point out that those limitations were mostly due to the neglected use of Latent State-Trait models and then to provide clear and empirical solutions. Thus, throughout the dissertation I have provided an overview of the current literature regarding the state-trait decomposition of GSE, have widely explained the above-mentioned limitations of previous literature, and have offered different methodological solutions, in order to (a) fill the gaps of previous literature, (b) enhance researchers’ capabilities to deal with similar unresolved conflicts, and (c) offer insights for future studies.