The need to belong and its association with fully satisfying relationships: A tale of two measures

Abstract

The current research compares the Need to Belong Scale (NTBS; Leary et al., 2013) and the Antecedents subscale of the Sense of Belongingness Inventory (SOBI-A; Hagerty & Putusky, 1995) to determine whether they represent approach or neuroticism-driven avoidance orientations in the need to belong. This research also extends previous research on these constructs to examine direct and moderating associations involving the need to belong and the quantity and quality of personal close relationships. Students (N=869) from a large university in the Southwest USA completed a battery of measures. Results indicated that the NTBS was associated with lower quality “partial” relationships rather than those of high quality “whole” relationships; this was not the case for the comparative SOBI-A. In addition, greater numbers of whole relationships buffered the effects of the NTBS on depression. The results are discussed in terms of the Belongingness Orientation Model.

Introduction

Research on belongingness has received extensive attention in the last 20 years. In their seminal paper that reviewed and extended the previous literature, Baumeister and Leary (1995) proposed that: “human beings have a pervasive drive to form and maintain at least a minimum quantity of lasting, positive, and significant interpersonal relationships” (p.497). These authors argued that lacking fulfillment of this need results in significant decrements to well-being. Though they surmised that the need to belong is fundamental, they also posited that there are individual differences associated with this need. Assuming that this is the case, it follows that people high in the need to belong should work harder to obtain satisfying relationships and may require a higher number of such relationships compared to individuals low in the need to belong. Moreover, such satisfying relationships should have a disproportionate effect on high need to belong individuals in fulfilling their belongingness needs and protecting them from feelings of negative affect such as depression and anxiety. This paper examines whether this is the case. We start with a review of two instruments proposed to assess this need: the Need to Belong Scale (NTBS; Leary, Kelly, Cottrell, & Schreindorfer, 2013) and the Sense of Belonging Instrument-Antecedents (SOBI-A; Hagerty & Patusky, 1995). We compare and contrast the measures—arguing that they tap different approach/avoidance motives (see Gable, 2006) in the need to belong, and thus should differentially relate to variables in the nomological net of belongingness-related variables. We then develop hypotheses regarding expected associations between the need to belong and satisfying relationships.

The NTBS was designed to assess individual differences in “the strength of the desire for acceptance and belonging” (p.3), and is the most well-known and used measure of the construct. Leary et al. (2013) argued that individuals scoring high in the need to belong regularly worry about acceptance and belonging. As a result, these concerns of being accepted and belonging lead them to “seek a large number of relationships, worry about how they are valued by others, and put a great deal of effort into sustaining interpersonal relationships” (p.3). This argument alludes to both positive and negative aspects of obtaining and maintaining relationships that are featured in the NTBS. With respect to items assessing the positive aspects of the construct (e.g., “I need to feel that there are people I can turn to in times of need”), the need to belong arguably motivates individuals to behave in ways that yield larger numbers of quality relationships to meet one’s relational needs. With respect to items assessing the negative aspects, the instrument also consists of items that imply difficulties in meeting one’s relational needs (e.g., “It bothers me a great deal when I am not included in other people’s plans”)—consistent with an insecure attachment style. Accordingly, Leary and colleagues used a large set of studies to provide evidence that their measure taps into both the positive and negative aspects of the need to belong. For instance, they found that the NTBS has positive associations with the need for affiliation, affiliation motivation, sociability, agreeableness, preference for affiliation, the degree to which individuals regard their social identity as important, and the degree to which individuals value secure and satisfying interpersonal relationships. Yet, the NTBS also showed positive associations with neuroticism, anxious attachment, depression (in one sample, but not another), fear of criticism and rejection, social isolation, shyness, embarrassability, propensity for hurt feelings, and dependent, avoidant, and borderline personality disorders. The latter associations suggest an avoidance orientation where relationships are sought to avoid a negative end state.

In a separate line of research, Hagerty and Patusky (1995) set out to develop an instrument to assess aspects of belongingness. Based on a factor analysis, their items divided into two factors that the authors termed sense of belonging (SOBI-P; Sense of Belonging Instrument-Psychological Experiences) and antecedents to belonging (SOBI-A; Sense of Belonging Instrument-Antecedents). The former measure was designed to assess achieved belonging in terms of valued involvement and person-environment fit. The latter measure, the SOBI-A, assesses the motivation and ability to belong. It not only includes items that assess the desire for belongingness interactions (e.g., “It is important to me that I fit somewhere in this world”), but also items assessing the degree to which one feels prepared to contribute successfully to close relationships (e.g., “Other people recognize my strengths and good points”). Thus, the item content suggests that the SOBI-A represents a more approach-focused version of the need to belong, and Hagarty and Patusky found that it correlates positively with achieved belonging, r = .45. That said, the nomological net for the SOBI-A has not been well established in previous research.

Accordingly, the first objective of the current study is to compare and contrast the correlational relations of the SOBI-A against those of the NTBS. Consistent with an approach orientation, we expect that the SOBI-A should correlate positively with achieved belongingness, positively valanced dimensions of the Big Five, and positively valanced measures of well-being (e.g., self-esteem), whereas it should correlate negatively with loneliness, anxious/avoidant attachment styles, and other negatively valanced assessments of personality and well-being. In contrast, we expect the NTBS to correlate positively with neuroticism and display associations consistent with an avoidance orientation, thereby replicating patterns obtained by Leary et al. (2013).

In addition, where previous research has examined the correlates of the NTBS with various trait-like measures and assessments of social support, there are no studies examining how the need to belong is explicitly associated with the quantity and quality of one’s close relationships. Our second objective is to examine this issue. To do so, we used criteria designed by Malone, Pillow, and Fuhrman (2013; manuscript in revision) to assess each close relationship reported by participants. Specifically, these criteria assess whether each close relationship is fully satisfying or not. The overarching criteria for defining a fully satisfying relationship are defined in the belongingness hypothesis advanced by Baumeister and Leary (1995), and are summarized as follows: “Satisfying this drive [to belong] involves two criteria: First, there is a need for frequent affectively pleasant interactions with a few people, and, second, these interactions must take place in the context of a temporally stable and enduring framework of affective concern for each other’s welfare” (p. 497). Based on Baumeister and Leary’s claims, Malone and colleagues constructed seven specific criteria to define fully satisfying relationships—termed here as “whole” relationships. Relationships that fail to meet all the criteria are termed “partial” relationships as they are only partially satisfying.

As such, this study will examine the associations of the NTBS and the SOBI-A with participant reports of whole versus partial close relationships. Assuming that the NTBS taps an avoidance orientation, we expect those scoring high on the NTBS to report fewer numbers of whole relationships and greater numbers of partial relationships. We further expect the SOBI-A to primarily assess an approach orientation in the need to belong, and hence expect that individuals scoring high on the SOBI-A to report greater quantities of whole relationships and fewer quantities of partial relationships. We further expect the implied correlations noted above to be stronger when using partial correlations to isolate the prevailing orientations of the NTBS and the SOBI-A (i.e., partial correlations controlling for the SOBI-A when assessing the prevailing deficit-reduction orientation of the NTBS, and partial correlations controlling for the NTBS when assessing the prevailing growth orientation of the SOBI-A).

Finally, we expect that whole and partial relationships may moderate the positive association between the need to belong and depression. Specifically, greater numbers of whole relationships should at least partially satisfy the need to belong and thus reduce levels of depression.

Method

Participants

The sample consisted of 869 undergraduate students from a large university in the Southwest USA who participated in Fall, 2010 (see Table 1 for socio-demographics). The original sample consisted of 875 participants,¹ but 6 participants were not included because they had missing data. The Institutional Review Board approved the study and participants received partial course credit for one hour of participation.

Table 1.

Sociodemographic Information

	%
Gender
Male	38
Female	62
Ethnicity
Hispanic	41
Caucasian	35
African-American	10
Asian	10
Other	4
Age (years)
16 – 17	2
18 – 19	73
20 – 22	15
23 – 26	4
27 – 55	3
Missing	3

Measures

Higher scores for each of the measures described below indicate a greater propensity for each scale consistent with the connotations in the labeling. Coefficient alphas, where applicable ranged from .76 to .96.

Need to Belong Scale (NTBS)

This 10-item measure, rated on a 5-point Likert scale, assesses the motivation to be accepted by others and avoid being shunned (Leary et al., 2013).

Sense of Belonging Instrument-Antecedents (SOBI-A)

This 14-item measure, rated on a 4-point Likert scale, assesses the potential and energy for meaningful involvement with others (Hagerty & Patusky, 1995).

Sense of Belonging Instrument-Psychological Experiences (SOBI-P)

This 18-item measure, rated on a 4-point Likert scale, assesses a general sense of achieved belonging. (Hagerty & Patusky, 1995).

Short Version of the UCLA Loneliness Scale (ULS-8)

This 8-item measure, rated on a 4-point Likert scale, assesses how much a person feels cut off or separated from others. These feelings are assessed by the lack of social contact one possesses compared to what is desired (Hayes & DiMatteo, 1987).

Satisfaction with Life Scale

This 5-item measure, rated on a 7-point Likert scale, assesses global satisfaction with life (Diener, Emmons, Larsen, & Griffin, 1985).

Subjective Happiness Scale

This 4-item measure, rated on a 7-point Likert scale, assesses global subjective happiness (Lyubomisrky & Lepper, 1999).

Center for Epidemiologic Studies Depression Scale (CES-D)

This 20-item measure uses a 0 to 3 scoring system to assess depressive symptomatology in the general public (Radloff, 1977).

Social Assurance Scale

This 8-item measure, rated on a 6-point Likert scale, assesses needs of reassurance from one or more persons. Lower needs indicate confidence in social situations (Lee & Robbins, 1995).

Experiences in Close Relationships-Short Form (ECR-S)

This 12-item measure, rated on a 7-point Likert scale, assesses avoidance and anxious adult attachment dimensions in intimate relationships (Wei, Russell, Mallinckrodt, & Vogel, 2007).

Self-esteem Scale

This 10-item measure, rated on a 4-point Likert scale, assesses an individual’s perception of global self-worth (Rosenberg, 1965).

Ten-Item Personality Inventory (TIPI)

This 10-item measure, rated on a 7-point Likert scale, assesses the Big Five constructs (Extraversion, Neuroticism, Agreeableness, Conscientiousness, and Openness to Experience) using one positive and one negative item per dimension. (Gosling, Rentfrow, & Swann, 2003).

Assessment of whole versus partial relationships

Participants were asked to make a list of 1 to 14 people with whom they perceived themselves as having a very close relationship. Subsequently, for each person on the list participants were asked to provide information regarding their relationship with that person. This information included seven questions and rating scales that assessed the criteria for defining a fully satisfying relationship. As described more fully in Malone et al. (2013), these criteria were derived from the belongingness hypothesis as put forth by Baumeister and Leary (1995). For a relationship to be classified as fully satisfying, or whole, the cut points across the entire set of criteria must be met or exceeded. Relationships that did not meet or exceed the cut points across all seven criteria were classified as partial relationships. Accordingly, the current research defined fully satisfying relationships using the following seven criteria and cut points: frequency of interactions were once per week or more; agree that interactions were positive; negative interactions occurred no more than somewhat often; agree that the relationship has been stable over time; believe that the relationship will endure for the next 10 to 20 years or longer; agree that this person is concerned for me; and agree that this person has as much concern for me as I have for them.

Results

Do the NTBS and the SOBI-A represent different orientations in the need to belong?

Before answering this question, we first established that the NTBS and the SOBI-A are only moderately correlated, r=.45, p<.001, yet strong enough to remain consistent with the development of both instruments as markers of the need to belong. Next, we demonstrated that they differ substantially in how they relate to other variables. For instance, the NTBS correlated .34 with loneliness whereas the SOBI-A correlated −.12 with the same measure. This pattern of differential directional effects from positively correlated measures suggests a suppression effect. Consistent with a suppression effect, the NTBS more strongly correlates with loneliness (r=.44) after controlling for the variance shared with the SOBI-A. This confirms that the NTBS may be best construed as an avoidance-focused measure. In contrast, the SOBI-A correlated −.32 with loneliness after controlling for the NTBS, thus indicating that the SOBI-A functions substantially as an approach-oriented measure of the need to belong.

Strong evidence of differentiation in the constructs is shown in the patterns of correlations provided in Table 2. Of the 14 zero-order correlates examined, the NTBS and SOBI-A differed in direction with respect to 11 of the correlates. Here, differences related to attachment style, self-esteem, and personality stand out. Specifically, the NTBS was positively related to anxious attachment style, low self-esteem, and neuroticism—consistent with an avoidance orientation; whereas the SOBI-A was negatively related to avoidant attachment style, and was positively related to self-esteem and extraversion—consistent with an approach orientation. As shown in Table 2, the pattern of differentiation is even stronger as evidenced by the partial correlations. Partial correlations were utilized to provide symmetric assessments of association and to isolate the two orientations of the two measures by controlling for their shared variance.

Table 2.

Differential Correlation Patterns for the Need to Belong Scale and Sense of Belonging Instrument-Antecedents.

Variables	Zero-Order Correlations		Partial Correlations
	NTBS	SOBI-A	NTBS	SOBI-A
Predictive Validity
Achieved Belonging	−.27**	.22**	−.42**	.39**
Loneliness	.34**	−.12**	.44**	−.32**
Depression	.31**	−.11**	.41**	−.30**
Life Satisfaction	−.19**	.15**	−.30**	.27**
Happiness	−.23**	.18**	−.37**	.33**
Attachment Orientation and Self-views
Social Assurance	.49**	.26**	.43**	.05
ECR-S Avoidance	−.04	−.26**	.10**	−.27**
ECR-S Anxious	.53**	.12**	.53**	−.15**
Self-esteem	−.36**	.13**	−.48**	.35**
Personality Dimensions
Extraversion	−.08*	.23**	−.22**	.30**
Neuroticism	.35**	−.03	.41**	−.23**
Agreeableness	.01	.26**	−.09**	.23**
Conscientiousness	−.07**	.21**	−.20**	.29**
Openness	−.09*	.24**	−.22**	.31**

Note. NTBS=Need to Belong Scale; SOBI-A=Sense of Belonging-Antecedents; ECR-S=Experience in Close Relationships-Short Form. Partial correlations for the NTBS controlled for the SOBI-A, and partial correlations for the SOBI-A controlled for the NTBS. Neuroticism is the reverse-scored measure of Emotional Stability.

^*p<.05,

^**p<.01.

How does the need to belong relate to quantities of high quality close relationships (wholes) versus relationships that are not considered fully satisfying (partials)?

We expected the answer to depend on whether the need to belong follows an approach or avoidance orientation. Hence, we first examined the NTBS and SOBI-A as commonly measured, and then re-evaluated the correlations using residualized NTBS and SOBI-A measures to better isolate approach versus avoidance orientations. The residualized NTBS scores were obtained by regressing the NTBS onto the SOBI-A, and then saving out the residuals—thereby capturing the variance in the NTBS controlling for the SOBI-A. To obtain the residualized SOBI-A scores, the predictor and criterion were reversed.

First, we found that the NTBS did not correlate with the total number of close relationships listed by participants, r=.03. However, when these relationships were divided into quantities of whole versus partial relationships, a small but significant correlation between the NTBS and quantities of partial relationships was obtained, r=.11, p<.01. On the other hand, the NTBS correlated at −.06 with quantities of whole relationships, and this association was not statistically significant, p=.10. The same pattern of correlations was obtained using the residualized NTBS to control for the approach orientation represented by the SOBI-A. Here the NTBS correlated −.10 with whole relationships, and .08 with partial relationships, ps<.05. Using a Fisher’s Z test for non-independent samples, we found the difference between the correlations of the residualized NTBS with whole and partial relationships to be statistically significant, Z= −3.85, p<.001. This pattern indicates that the NTBS is differentially related to quantities of whole versus partial relationships in such a way that is consistent with an avoidance orientation.

In contrast, the SOBI-A correlated .11 with total numbers of relationships, .08 with whole relationships, and .08 with partial relationships, all ps<.05. In addition, the residualized SOBI-A correlated .11 with total number of relationships and .12 with whole relationships, ps<.01. However, its correlation with partial relationships was near zero and not significant, r=.04; this correlation differed from the correlation of .12 at only a marginal level via a Fisher’s Z-test, Z=1.69, p=.09. Nonetheless, the overall pattern of results is consistent with an approach orientation in the need to belong, and indicates that those with such an orientation tend to establish more close relationships.

Do quantities of whole relationships buffer the depression-related consequences of a high need to belong?

To examine this question, a regression analysis was conducted with depression serving as the dependent measure. The residualized NTBS and whole relationships served as predictors. The former measure controlled for the SOBI-A to yield an avoidance-oriented assessment of the need to belong. These predictors were centered and then multiplied to form an interaction term per Aiken and West (1991). The main effects obtained were statistically significant (for the residualized NTBS, β=.387, t(865)=12.67, p<.001; for wholes, β=−.18, t(865)= −5.95, p<.001), but were qualified by the hypothesized 2-way interaction. As expected the interaction between the NTBS and quantities of whole relationships was found to be statistically significant, β=−.037, t(865)= −2.68, p<.01, though the effect was small, ΔR²=.007. The interaction was decomposed into simple slopes as shown in Figure 1. Here it was found that individuals reported higher levels of depression as the need to belong increased, and these simple slopes were statistically significant for those low in quantities of whole relationships (i.e., those 1 SD below the mean), β=.47, t=10.63, p<.001, as well as for individuals high in quantities of whole relationships (i.e., those 1 SD above the mean), β=.29, t=6.33, p<.001. That said, higher quantities of whole relationships did buffer the effects of need to belong on depression—though they did not appear to fully satisfy the need to belong—a result that is consistent with an avoidance orientation. A similar regression analysis was conducted using the residualized SOBI-A in place of the residualized NTBS, but the interaction effect was not found to be statistically significant.

Figure 1.

This graph shows that positive relation between the NTBS and depression was partially buffered for individuals with high quantities of whole relationships. The graph displays unstandardized slopes.

Discussion

How should we construe the need to belong as measured by Leary et al. (2013)? The data presented here indicate that the NTBS largely assesses a neuroticism-driven, avoidance orientation with respect to the need to belong. Evidence for this conclusion comes in three forms. First, we found that the NTBS correlated highly with a constellation of variables that represent negative self-views, negative working models of relationships, preoccupation with acceptance from others, worry, and anxiety—especially after controlling shared variance with the SOBI-A. These results are generally consistent with Leary et al.’s findings, though they found relatively more positively valenced correlates of the NTBS than did we. This constellation of variables suggests that the NTBS largely assesses a state stemming from one’s belongingness needs going unmet. In stark contrast, the SOBI-A—which also assesses the need to belong—was found to correlate positively with variables representing well-being (e.g., belongingness, happiness, self-esteem, and extraversion).

Secondly, we found that individuals who scored high on the NTBS tend to report having more partial relationships (i.e., partially satisfying) than whole relationships (i.e., fully satisfying), whereas individuals scoring high on the SOBI-A reported more whole relationships. There are a number of reasons why this pattern may result. For instance, individuals high on the NTBS may interpret ambiguous behaviors as ignoring or rejecting; they may experience increased negative affect, resulting in decreased opportunities to obtain and maintain fully satisfying relationships; or, they may exhibit overly dependent patterns of behavior that lead others to back away. Consequently, their concerted efforts to achieve belongingness may be creating a cycle that instead thwarts those efforts.

Finally, we found that high numbers of fully satisfying relationships partially buffered the relationship between the NTBS and depression after controlling for the approach-oriented SOBI-A. We reasoned that whole relationships should fully buffer against depression for those high on the NTBS if their need to belong could be fully satisfied; however, a fully buffering effect was not found. Instead, the partial buffering effect is consistent with that idea that it is difficult for persons high on the NTBS to find relationships that are completely satisfying.

Interestingly, our results are consistent with the Belongingness Orientation Model (BOM) advanced by Lavigne, Vallerand, and Crevier-Braud (2011) who found that the need to belong can be described with respect to two orientations that differ in valence: a growth orientation (akin to an approach-focused motive) versus a deficit-reduction orientation (akin to an avoidance-focused motive). The growth orientation refers to a fundamental interest in obtaining satisfying interpersonal relationships. Consistent with our results, Lavigne and colleagues found that this orientation was positively associated with a secure attachment style. Conversely, those with a deficit-reduction orientation “desire [the] closeness of others to fill a social void [and constantly crave] social acceptance” (p. 1186). Lavigne et al. found that this orientation was correlated with preoccupied and fearful avoidant attachment styles. Similarly, we found that NTBS was correlated with insecure attachment. These results are consistent with a tendency to react poorly to others’ reluctance to get close and may reflect behavior patterns (e.g., trying too hard to gain approval) that can lead to rejection from others. Additionally, Lavigne et al. found that that the NTBS was strongly and positively correlated with their own measure of deficit-reduction orientation (r=.59) and was only weakly associated with their measure of a growth orientation (r=.08, n.s.). Also, the partially buffering interaction is consistent Lavigne et al.’s (2011) argument that those with a deficit-reduction orientation experience difficulties in meeting their relationship needs, hence leading them to constantly desire attention and reassurance.

Lavigne et al.’s (2011) proposed growth and deficit-reduction orientations to achieve a sense of belonging build nicely on work advancing more general approach versus avoidance orientations that have been previously examined in relationship research and in other areas (e.g., hope for affiliation versus fear of rejection and the need to achieve success versus avoid failure), and has a long and prolific history (e.g., Lewin, Dembo, Festinger, & Sears, 1944; McClelland, Atkinson, Clark, & Lowell, 1953). Positively and negatively valenced outcomes for approach versus avoidance motivations have often been demonstrated. For instance, contemporary research by Gable (2006) showed that approach motives were associated with less loneliness and more satisfaction with social bonds, whereas avoidance motives were associated with more loneliness, negative social attitudes, and relationship insecurity. Indeed, the research provided in this study provided evidence consistent with Levigne et al. that the NTBS assesses a deficit-reduction orientation and is associated with increased loneliness and less relationship security, whereas the SOBI-A is associated with less loneliness greater relationship security, less neuroticism, and greater extraversion—all consistent approach motivations or a growth orientation to belongingness.

It is important to note that the present research did not set out to fully define the need to belong as a construct, only to shed light on features of the construct via contrasting two measures designed to assess it. That is, by gaining a better understanding of the approach-avoidance potentials associated with the need to belong we can provide a more comprehensive body of work. For example, though Leary et al. (2013) showed that the NTBS had a positive correlation with the importance that people place on close relationships with friends, partners, and family members, the research presented here shows that individuals high on the NTBS may have issues in obtaining and maintaining such close relationships—that is, high scores on the NTBS showed a negative association with numbers of whole relationships and a positive association with numbers of partial relationships, and the difference between those correlations was significant.

Limitations

The current study has its limitations. First, the study was conducted with university students, and the nature of one’s need to belong, as well as the nature of one’s close relationships, may differ in important ways for persons in their late teens and early twenties compared to older adults. Second, the design of the current study was cross-sectional and correlational, and it would be advantageous to show that the interaction effect obtained can be replicated in intervention study that designed to improve the quality of social connections. Third and finally, whole and partial relationships were demarcated using specific cutpoints and criteria that remain under investigation. Changing either the criteria, the cutpoints, or using other operationalizations of high quality versus unsatisfying relationships have the potential to adjust the estimates of effects presented here. Further research using diverse samples, various methodologies, and alternate operationalizations of fully satisfying versus partially satisfying relationships is needed to address these issues.

Conclusion

This is one of the first studies to expand the nomological net of the SOBI-A as it has not been widely used by psychologists. This study also adds to the literature using a comparative approach to further define the nature of the NTBS. This is also the first study to show that high quality relationships that involve frequent interactions are more common among those scoring high on an approach-focused assessment of the need to belong (i.e., the SOBI-A), and the first to show that those scoring high on an avoidance-focused measure of the need to belong (i.e., the NTBS) have more partially satisfying relationships. This research may in the future prove useful in helping clinicians better assess those high in the need to belong, and in helping those with an avoidance orientation to understand the potential aversive effects of this need and adjust or adapt their feelings and needs accordingly in an effort to obtain and maintain the close relationships they so highly value. Further work is needed on both theoretical and practical grounds to develop strategies that lead individuals to fully buffer against the pernicious effects of this avoidance-focused need. Finally, further work is needed to integrate our findings with the Belongingness Orientation Model.

Highlights.

• We examine the correlates of two instruments that assess the need to belong.

• These instruments differ in whether they tap approach or avoidance orientations.

• The Need to Belong Scale is associated with less fully satisfying relationships.

• Fully satisfying relationships partially buffers against predicted depression.