Optimal human functioning necessarily requires a relationship with others (Ryan and Deci, 2017). Relational needs have been strongly associated with improved health (vitality, self-determination, self-control, …) and wellbeing (Maslow, 1987; Baumeister and Leary, 1995; Kasser and Ryan, 1999; La Guardia et al., 2000; Ryan and Deci, 2000) and despite differences in theoretical approaches, there is a remarkable convergence among scientists on the fundamental importance of being connected to others (La Guardia and Patrick, 2008; Ryan and Deci, 2017).

Within the specific framework of integrative relational psychotherapy and transactional analysis, the relational needs model developed by Erskine (1996, 1997, 1998) is widely recognized (Pourová et al., 2020). This model emerged from a study of transference in psychotherapy and a qualitative investigation of the crucial factors in significant relationships conducted at the Institute for Integrative Psychotherapy in New York City in the early 1990s. However, relational needs are not only present in the context of psychotherapy; they are essential to a person’s sense of wellbeing throughout the life cycle (Erskine, 2015). In this respect, relational needs are the component parts of a universal human desire for the relationship; they are the needs unique to interpersonal contact (Erskine, 2015, p. 46). Relational needs are not the basic needs of life, such as food, air, or proper temperature, but they are the essential elements that enhance the quality of life and a sense of self-in-relationship. When a relational need is not satisfied, the need becomes more intense and is phenomenologically experienced as longing, emptiness, nagging loneliness, or an intense urge often accompanied by nervousness (Erskine et al., 1999). The continued absence of satisfaction with relational needs may be manifested as frustration, aggression, or anger. When disruptions in the relationship are prolonged, the lack of need satisfaction is manifested as a loss of energy or hope and shows up in script beliefs, such as “No one is there for me” (Erskine et al., 1999, Erskine and Moursund, 2011). These script beliefs are the cognitive defenses when needs do not get a satisfying response from another person (Erskine, 1980).

Žvelc et al. (2020) have emphasized this need for the relationship as a primary human motivation and have substantiated their perspective on several child development researchers and writers (Fairbairn, 1952; Bowlby, 1969, 1973, 1980; Kohut, 1971; Ainsworth et al., 1978; Stern, 1985; Fairbairn, 1986/1941; Winnicot, 1986/1960; Hesse, 1999). Attachment systems motivate infants to seek proximity and communication with caregivers. Therefore, attachment has an important evolutionary function—to heighten the possibility of survival of the child. Additionally, attachment relationship is crucial for the healthy development of both the brain and interpersonal relationships (Schore, 1994, 2001, 2003; Siegel, 1999; Cozolino, 2002). The qualities of affect and rhythmic attunement, relational needs, and sustained gestures of attachment between child and parent that provide the regulation of a child’s body sensations and affects are crucial for establishing a sense of safety, connection, and secure attachment (Erskine, 2021). In fact, attachment theory and research have underscored the importance of attachment and relational needs through the life cycle (Hazan and Shaver, 1987; Bartholomew and Horowitz, 1991; Hesse, 1999; Wallin, 2007).

The relational needs identified by Erskine and included in his model are the following (Erskine et al., 1999; Erskine, 2021): (1) security is the visceral experience of having our physical and emotional vulnerabilities protected. It involves the experience that our variety of needs and feelings are human and natural. Security is a sense of simultaneously being vulnerable and in harmony with another; (2) to feel validated, affirmed, and significant within a relationship requires the other person’s validation of the significance and function of our intrapsychic processes of affect, fantasy, and meaning-making, and to validate that our emotions are a significant intrapsychic and interpersonal communication. It includes the need to have all of our relational needs affirmed and accepted as natural. This need is a relational request for the other person to be involved by providing a quality of interpersonal contact that validates the legitimacy of relational needs, the significance of affect, and the function of intrapsychic processes; (3) acceptance by a stable, dependable, and protective other person is an essential relational need. Each of us as children had the need to look up to and rely on our parents, elders, teachers, and mentors. We need to have significant others from whom we gain protection, encouragement, and information. The relational need for acceptance by a consistent, reliable, and dependable other person is the search for protection and guidance; (4) the confirmation of personal experience is also an essential relational need. The need to have experience confirmed is manifested through the desire to be in the presence of someone who is similar, who understands because he or she has had a similar experience, and whose shared experience is confirmed. It is the quest for mutuality, a sense of walking, together with a companion who is “like me,” who has walked the same path in life. It is the need to have someone appreciate and value our experience because they phenomenologically know what that experience is like; (5) self-definition is the relational need to know and express one’s own uniqueness and to receive acknowledgment and acceptance from the other. Self-definition is the communication of one’s self-chosen identity through the expression of preferences, interests, and ideas without humiliation or rejection; (6) another essential relational need is to have an impact on the other person. Impact refers to having an influence that change the other in some desired way. An individual’s sense of competency in a relationship emerges from agency and efficacy, attracting the other’s attention and interest, influencing what may be of interest to the other person, and affecting a change in affect or behavior in the other; (7) the need to have the other person initiate refers to the impetus of another person’s making interpersonal contact. It is the reaching out to another, in some way that acknowledges and validates their investment in the relationship; and (8) the need to express love is an important component of relationships. Love is often expressed through quiet gratitude, thankfulness, giving affection, or doing something for the other person. The relational need to express love, whether it be from children to parents, sibling or teacher, or from a client to a therapist, is an important component in maintaining relationships. When the expression of love is stymied, the expression of self-in-relationship is thwarted.

Žvelc and Jovanoska (2016, 2017) and Žvelc et al. (2020) developed a new instrument for measuring the satisfaction of relational needs in the general population (RNSS). The final version is comprised of 20 items rated on a five-point Likert scale from agree to disagree. The scale yields five subscales and a total score of satisfaction of relational needs. The first four subscales (support and protection, have an impact, shared experience, and initiative from the others) reflect the relational needs described by Erskine. According to Žvelc et al. (2020), the fifth scale reflected the other four dimensions of the model as an overall dimension and was named need for authenticity. The instrument has been validated in Slovenian (Žvelc et al., 2020), Czech (Pourová et al., 2020), and Turkish (Toksoy et al., 2020) and showed good psychometric properties and the confirmation of the factorial structure and the hierarchical model.

Given the lack of instruments to measure relational needs in Spanish, the purpose of this article is to adapt the Relational Needs Satisfaction Scale (RNSS) using both a clinical and non-clinical sample and to analyze its psychometric properties. It is expected that the Spanish version of the RNSS presents adequate psychometric properties in terms of reliability and construct and discriminant validity, thus offering new evidence of the validity of the instrument.

The main objective of this study was to adapt to Spanish a scale that allows the measurement of relational needs for subsequent use in clinical practice contexts and in research with non-clinical populations. The observed results allow us to conclude that the Spanish adaptation of the RNSS offers adequate characteristics of language adaptation and measurement.

Inter-rater agreement of the translated version was appropriate. Likewise, the assessment made by 10 people of the characteristics of the item statements allows us to accept that they are easy to understand, that the items are related to the construct they intend to measure, and that the items are in the repertoire of interpersonal relationships (see Supplementary Table 1 for the final version of the Spanish scale). However, in the pilot study, it was already possible to identify some items that showed a differential behavior. Of these, item 18 is the item with the greatest number of problems in the psychometric adequacy of the scale. Nevertheless, we believe that the results obtained in the pilot study provide evidence for the face and content validity of the RNSS and that the translation is appropriate.

In general, the metric behavior of the items is acceptable. Mostly negative asymmetries are observed, indicating a tendency on the part of the participants to report higher relational satisfaction scores. This effect is more marked in the items belonging to the “protection” dimension (items 04 and 13), but there are no outliers in any of the remaining items. In this respect, item response behavior follows a pattern similar to that observed in other instruments used in psychological assessment (Holgado-Tello, 2015).

With respect to the internal consistency of the items, item 18, integrated into the dimension of “initiative from others,” performed poorly. This item is the only one stated negatively and, as mentioned previously, it was one of the items that presented the greatest difficulties of comprehension. It is very likely that its different wording produces a response bias that could perhaps be corrected by changing its wording to positive.

Independently of the effect observed in item 18, the internal consistency shown for the total scale was high (α = 0.87), very similar to that obtained in the original study (Slovenia, α = 0.90), and in the two countries that adapted it to their language (the Czechia, α = 0.90; and Turkey, α = 0.83). However, if we look at the reliability observed in the dimensions, we see that performance is lower in the present study than in the Slovenian or Czech studies (Pourová et al., 2020; Žvelc et al., 2020), and more similar to the results obtained in the Turkish study (Toksoy et al., 2020).

In our study, the internal consistency observed in the community and clinical sample was equivalent except for the authenticity dimension. It has also been observed that, with the exception of the dimensions initiative from others and having an impact, the clinical sample had a statistically significant lower perception of satisfaction of needs in their relationships. This result allows us to accept that the RNSS shows evidence of discriminant validity.

Another important aim of our study was to confirm the factor structure of the RNSS. We have found empirical support that allows us to accept both the hierarchical structure of five factors subsumed in an overall factor proposed in the original study (Žvelc et al., 2020) and in the two adaptation studies (Pourová et al., 2020; Toksoy et al., 2020), as well as the proposed bi-factor model. Of the two, we favor the bi-factor (Bi-F) model based on the following criteria. Several authors (Reise et al., 2010, 2013; Chen et al., 2013) have summarized the advantages of the Bi-F model over second-order confirmatory factor analysis (SOCFA). The main advantage is that the Bi-F analysis allows to directly observe to what extent an item or scale (the observed variable) reflects a common target construct (i.e., a general factor) and, simultaneously, to what extent it may reflect a (domain-specific) sub-dimension. Consequently, the Bi-F model allows for the retention of a single common latent factor but also controls for variance arising due to additional common factors.

A second advantage of the Bi-F model is that in the SOCFA model it is not possible to observe direct relationships between the observed variables and the general factor, but rather an “indirect effect” or a “mediated relationship” through the first-order factors. Therefore, to estimate the variance attributable to the general factor, the loading of the observed variable on the domain-specific factor must be multiplied by the loading of the domain-specific factor on the general factor (Chen et al., 2006, 2013). In contrast, the Bi-F model provides information on all factor loadings and makes it possible to identify whether a domain-specific factor uniquely contributes to the prediction of external criteria (Chen et al., 2006, 2013). Given that in a Bi-F model the general and domain-specific factors are orthogonal, even a simple inspection of the item loadings on the second- and first-order factors is informative.

Since the SOCFA model is nested in the Bi-F model, no such restrictions are present. The latter can be used as a base model to compare model fit as the model becomes more constrained, which is a third advantage of the Bi-F model over SOCFA (Brown, 2014). For example, in a SOCFA model, correlations between first-order factors are assumed to occur because they have a common cause (i.e., the overall factor). Therefore, the observation of low loadings on a domain-specific factor and high loadings of the specific factors on a general factor may suggest that these variables are best explained by a general factor and do not constitute a domain-specific factor. Thus, if the items mainly reflect the general factor and have low loadings on the first-order factors, the sub-scales make little sense.

Our results with the SOCFA/hierarchical model are very similar to those obtained in the previous studies (Pourová et al., 2020; Toksoy et al., 2020; Žvelc et al., 2020) and, as the researchers have concluded, the five-dimensional model that would represent Erskine’s theory and the option of a general factor that would allude to the concept of relational needs would be satisfied. However, the Bi-factor model offers other possible conclusions. The data collected in the Spanish sample preferably offered the option of a general factor. The Velicer analysis and the solution of the Bi-factor model seem to favor this option. The factor loadings are of larger effect size than in the specific dimensions, although this does not rule out the contribution of the items to the latent factors representing the dimensions explored. Further work is needed to confirm this solution, and one possibility would be to seek an integration of all the data obtained in the various studies that have examined the evidence for the validity of the RNSS.

In sum, the adaptation to Spanish of the RNSS has shown an appropriate inter-rater agreement in the translation of the Spanish statements, as well as in the functionality of the item comprehension and the use of the relational needs valuation. In other words, the scale shows face and content validity. Besides, the results demonstrated that the RNSS has an adequate accuracy of measurement with sufficiently high coefficients and appropriate construct validity that is expressed through a suitable index of adjustment on the confirmatory factorial analysis. Therefore, this study gathers evidence to consider the translated RNSS as an appropriate instrument for the measurement of the mentioned construct. Nevertheless, it is necessary to adjust some aspects of the formulation of an item (number 18) and the execution of new studies with more controlled designs to demonstrate other evidence of validity. It will be advisable to check the invariance of the dimensional structure of the RNSS in different populations (clinical vs. community samples) or in different cultures and relational contexts (individualistic vs. Collectivistic cultures). Furthermore, we could also test the discriminative power of the RNSS in these conditions and the predictive validity as it regards functionality, health, and wellbeing. Finally, the instrument has been developed after Erskine’s theory of relational needs, and thus it can be a useful tool within the framework of integrative psychotherapy, both for diagnostic and assessment purposes.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethical Committee of the University of Deusto (ETK-19/19-20). The patients/participants provided their written informed consent to participate in this study.

II designed the study, articulated the methodology and results in sections of the manuscript, conducted the data analysis, and wrote the draft manuscript. IG-M, AM, and SM-R played a role in item development and data collection. RE supervised the research and contributed to the writing of the manuscript. SG was responsible for writing the manuscript and contributed to the discussion of the results. GŽ contributed to the design of the study and the discussion. All authors reviewed and approved the final version of the manuscript.