External theoretical reference
Do Similar People Really Like Each Other More?
An accessible guide to outside research: the questions it addresses, the evidence behind it, and the limits of its application.
Plain-language question: Which is closer to the research: “birds of a feather flock together” or “opposites attract”?
Montoya and colleagues synthesized 313 laboratory and field studies comprising 460 effect sizes. Overall, both actual similarity and perceived similarity were associated with attraction. The stage of the relationship mattered, however. Actual similarity was more important in no-interaction and brief-interaction studies; its effect dropped markedly beyond brief interactions and was nonsignificant overall in existing relationships. By contrast, “I feel that we are alike” predicted attraction before interaction, during brief interaction, and in established relationships. S06: Montoya, Horton, & Kirchner, 2008
Where the evidence comes from: This is a large meta-analysis whose strength is comparing different stages of relationships. Its limitations are that much of the evidence is correlational and that perceived similarity may cause liking or may be an explanation people form after they already like someone. The studies also examine different kinds of similarity, including attitudes and personality; these cannot be collapsed into a single universal switch.
Illustrative example: Two people quickly find common ground when they first talk about films because their tastes overlap. Six months later, what sustains their closeness may be less the overlap in their watchlists than the shared feeling that the other understands their choices. Another pair may receive different test results yet see each other as “people who refuse to be shortchanged”; their sense of similarity is still real.
What this contributes to the project: “Mirroring” in M02 can mean a shared logic, blind spot, or story that both people recognize; their four letters need not match. A character can say, “Your methods differ from mine, but I know why you do it.” This comes closer to similarity’s relational role than awarding points for matching letters.
What it does not show: We cannot say that identical types are the best match, or that different types necessarily lack common ground. Matching four-letter codes mean only that the characters share strategic tendencies as defined by this project. They do not imply identical values, habits, warmth, or relationship goals.