What the Social Relations Model Separates in Friends’ Ratings

What does the model separate in a group of friends?

Suppose several friends describe one person as dependable, while one friend calls that person unreliable. Before deciding that the person behaves inconsistently, ask what the ratings show about the whole group. The Social Relations Model (SRM) is a way to separate parts of interpersonal judgments when the same people rate one another. In a round-robin design, each member rates every other member on a defined quality. That network supplies repeated views of the same targets and repeated ratings from the same perceivers.

The SRM distinguishes three components. The perceiver effect follows the rater: does this friend tend to rate people higher or lower, or see a particular quality in others? The target effect follows the person being rated: does this person receive similar ratings from several group members? The relationship effect is the remaining pair-specific difference: does this rater see this target differently than their general rating tendency and the target’s group-wide standing would lead us to expect? David Kenny’s methodological overview defines these components and describes the round-robin design. They are sources of variation in judgments, not three personality traits inside the target.

Think of the components as three questions about a rating. A generally generous rating style belongs to the perceiver component. A person repeatedly described as dependable belongs to the target component, which can be read as reputation in that group. One friend’s unusually low rating of that same person belongs to the relationship component, once the other two patterns are accounted for. A single friend-target score cannot establish all three patterns on its own; the separation depends on enough connected ratings to compare raters and targets.

Sources: Interpersonal Perception: Social Relations Model; What Do Peer Evaluations Represent? A Study of Rater Consensus and Target Personality

How is a rater’s general lens different from a shared reputation?

The perceiver effect is a pattern in what someone gives; the target effect is a pattern in what someone receives. If a friend calls nearly everyone thoughtful, that tendency changes how to read their rating of one person. If several friends independently describe that person as thoughtful, the shared impression follows the target. Both can raise a particular score, but they answer different questions.

A general rater tendency is not automatically a flaw to subtract away. In a 2021 study spanning ten studies, 2,199 perceivers, and more than 100,000 personality judgments, Rau and colleagues found that perceiver effects reflected both broad positivity or negativity and trait-specific tendencies. The balance varied across trait domains and judgment contexts. This suggests that one friend’s lens may be broad for one kind of judgment and more specific for another. It does not give us a correction factor for a particular person’s ratings.

A target effect is also narrower than an objective description. It summarizes how this sampled group sees someone, for the particular trait and rating task. Friends may share experiences, notice different settings, or hold common assumptions. So agreement supports the statement ‘this is a shared impression here.’ It does not alone support ‘this is the person’s true, complete personality.’ For a practical reading, check whether the same rater tends to use the scale generously across targets, then separately note whether multiple raters converge on the target.

Sources: Interpersonal Perception: Social Relations Model; What is the structure of perceiver effects? On the importance of global positivity and trait-specificity across personality domains and judgment contexts

What can a pair-specific rating mean, and what can’t it prove?

The relationship effect marks what is distinctive about one person’s view of one other person after their usual rating style and the target’s shared standing are considered. In the opening example, one friend’s unusually low dependability rating would be pair-specific if it sits apart from the friend’s ratings of others and the target’s ratings from the rest of the group. It says where the difference sits in the network. It does not name its cause.

That difference could arise because the two friends share particular situations that others do not see. It could reflect expectations, a real but limited pattern in their interactions, an ambiguous rating question, or simple imprecision. These are possibilities for interpreting a deviation, not findings the model can choose among. A 2022 peer-reviewed study of engineering student teams explains the residual relationship component as a distinctive perceiver-target rating and notes it can include unsystematic sources such as careless responding or misreading criteria. The authors studied skill ratings in structured teams, not broad personality judgments between friends, so their example clarifies the component’s limits rather than estimating how friend ratings usually divide.

This is why ‘unique’ should not be mistaken for ‘deep’ or ‘especially revealing.’ A pair-specific result is not proof of projection, conflict, closeness, hidden behavior, or a special bond. To explore it, ask what observable situations the two people share and whether the rating refers to a specific behavior. Then compare that account with other contexts and the target’s own perspective. The SRM isolates the difference for investigation; it does not settle the personal explanation.

Sources: Interpersonal Perception: Social Relations Model; What Do Peer Evaluations Represent? A Study of Rater Consensus and Target Personality

Does agreement show that friends have read someone accurately?

No. Agreement and accuracy are related but separate questions. Agreement asks whether raters converge. Accuracy asks whether a judgment corresponds to an appropriate criterion, such as observed behavior or a carefully chosen self-report, while recognizing that each criterion has limits. A target effect estimates shared standing among the sampled raters; it does not validate that standing by itself.

In three discussion-group studies reported by Paulhus and Reynolds, participants rated one another after meeting once and again after seven meetings. Rating validity improved with acquaintance, but target variance did not. In other words, judgments became more valid by the study’s criteria without the amount of between-target variation increasing. This is a useful demonstration that consensus and validity can move differently. Those were discussion groups with a particular design, not a general estimate of how accurately friends judge personality.

Context matters too. A study of families with adolescents used the SRM to examine Big Five judgments and found that patterns varied by trait and family role. The authors reported more consistent judgments for Agreeableness and Conscientiousness in that setting, alongside substantial relationship and perceiver effects. Family roles and shared history differ from friendship groups, so this finding supports caution about context, not a claim about which traits friends judge best.

The practical verdict: use the SRM to ask whether a difference belongs mainly to a rater’s usual lens, a group’s shared impression, or one pair’s distinctive view. Then treat each as a clue, not a verdict. If several friends agree, look for repeated observable behavior across settings; if one friend differs, ask what they uniquely observe. Keep the person’s own account in view, too. Those comparisons can deepen reflection, but they answer different questions and none turns a set of ratings into a complete personality portrait.

Sources: Interpersonal Perception: Social Relations Model; Enhancing target variance in personality impressions: highlighting the person in person perception; Personality judgements in adolescents' families: the perceiver, the target, their relationship, and the family

Questions readers ask