The Dimensional Structure of Human Relationships
The Dimensional Structure of Human Relationships
While interpersonal relationships are conventionally treated as explicit categories, I propose a model where such categories represent combinations of latent dimensions that underlie differences in interpersonal relationships. I conduct a study to reveal these underlying dimensions, then define different types of relationships as general combinations of dimensional traits, which fade into each other. The primary hypothesis in this study was that very close friendships and romantic relationships would overlap in a latent dimension of intimacy/closeness and differ in a dimension reflecting aspects of sexual attraction. Otherwise this study was primarily exploratory without initial predictions being made. A multidimensional model as opposed to conventional definitions of relationships is produced from the results.
Methods
An online survey was conducted using PsyToolkit (Stoet 2010)(Stoet 2017) where participants were asked to answer a set of questions about up to ten interpersonal relationships they have had. The total number of participants was small (n=27), however participants submitted responses for a total of 71 relationships.
Results
The relationship questionnaire included items reflecting characteristics about the relationship as well as their own categorization of and origin of the relationships. The questions assessing characteristics of the relationships only were principal components analysis, the scree plot suggesting extraction of 6 factors. These factors were not rotated as they were interpretable [Table 1].
Table 1
The first component had highest loadings on items reflecting fondness, attachment, loyalty, emotional/intellectual/physical/sexual intimacy, trust, and closeness and was called Intimacy/Closeness. The second component had negative loadings on items reflecting instability, conflict, control, and aspects of dissatisfaction with the relationship, and was called Healthy (vs toxic). The third component had positive loadings on items reflecting desire for intellectual intimacy and non-romantic or sexual activities, and negative loadings on items reflecting sexual and romantic attraction and intimacy, this factor was called Friendzone (vs fuckzone). The fourth component had strong loadings on items reflecting aspects of relationships that indicate the relationship was not equal or formed by mutual choice and was called Circumstantial (vs organic). The fifth factor had strong negative loadings on items reflecting diverging relationship desires and unreciprocated desires and feelings, and was called Requited (vs unrequited). The sixth factor had negative loadings on items reflecting a lack of desire for commitment or long term involvement with a person, and was called Committed (vs casual).
Correlations of the factor scores with self categorization of relationships were then calculated [Table 2]. The primary correlates defining each category are presented in table 3.
Table 2
Table 3
The primary factors correlating with the different categories of relationships were intimacy and friendzone-fuckzone, and the correlations of the categories (that are large enough to be relevant) are plotted in figure 1. A 45-degree rotation of the correlates reveals orthogonal dimensions of friendship and romance, where best friend loads highest on friendship and acquaintance lowest, and crush & significant other about equally high on romance and relatives lowest on romance.
Figure 1
The primary factors differentiating different types of sexual relationships were the commitment and requited dimensions, and the categories with notable correlations were plotted [figure 2].
Figure 2
Discussion
The primary hypothesis was supported. It was found that best friends, significant others, and crushes were all characterized by high levels of the intimacy/closeness factor, and best friends were characterized by desire for emotional and intellectual intimacy but not sexual or romantic involvement, significant others characterized by an equal desire for emotional/intellectual intimacy and sexual/romantic involvement, and crushes by lower desire for emotional/intellectual intimacy and a high desire for sexual/romantic involvement. These findings make sense, given that a crush is typically an attraction formed from a distance before one attempts to become involved with the person, where one would be fearful of revealing themselves emotionally/intellectually in case it were to end in rejection. Couple relationships are distinctly characterized by both emotional/intellectual intimacy and sexual/romantic involvement, therefore high in closeness but not having an imbalance to one form of intimacy. Close friendships are characterized by high intimacy but generally a lack of sexual/romantic involvement but high emotional/intellectual intimacy.
Different forms of relationships with sexual/romantic involvement were differentiated by the dimensions of commitment and requittedness. Significant others and open relationships were more required compared to friends with benefits and situationships, reflecting that couple relationships and open couple relationships tend to be more mutual and equally desired, whereas friends with benefits and situationships more often involve a discrepancy in what each person desires from the relationship. Significant others were differentiated from all other sexual relationships by the commitment dimension, which is expected given that open relationships, friends with benefits, and situationships are all characterized by a lack of commitment or being defined in terms of a couple.
The other dimensions largely reflected aspects that were not defining of the form of relationship, but rather reflected emergent and originating aspects of a relationship, respectively the degree to which it is healthy or unhealthy, and whether the relationship was formed through mutual engagement or by chance (e.g., a relative, boss, coworker).
Conclusion
Overall, this study provides an alternative non-categorical model for defining interpersonal relationships, which often do not fit precisely into conventional categories and can have unique blends of the characteristics defining these conventional categories.
Stoet (2010). PsyToolkit - A software package for programming psychological experiments using Linux.
Stoet (2017). PsyToolkit: A novel web-based method for running online questionnaires and reaction-time experiments
Comments
Post a Comment