Psychology 210: Social Psychology

R. Akert


Example of Kelley’s Model of Covariation: “The Woman who flirts…”


            You’re at a party and you observe a woman you know flirting (quite outrageously!) with a man you’ve seen around. This is very interesting to you, so you decide to think about it (and maybe discuss it with a friend?); you’re trying to figure out what’s going on. The attributional question (in a sense) that you’re asking yourself is:


            “Why is she acting this way (flirting) with him? What does it mean?”


            According to Kelley, to make a strong, confident attribution (and one hopes, an accurate one!), you need three kinds of information: Consistency, Consensus, and Distinctiveness. “Data” of these three types will help you decide if it is something about her (an INTERNAL attribution) that is causing her behavior or something about “the situation” (the guy; the party situation; the presence of other people; etc.) – an EXTERNAL attribution—that is causing her behavior. Let’s see how this might play out:


Consistency information:


            You need some “data” about her and how she interacts with this guy at other times, other places, etc. If you’ve observed her before with him (e.g., at past parties, other social settings), you pull this information out of memory and examine it. And/or, you could ask friends if they’ve observed these two together before. So, you wondering if her behavior is “consistent”—has she been flirty with him at other times and in other situations (like at a funeral!)?


            HIGH consistency = Yes, she’s flirted with him before (even at that funeral!). (Hmmmm…could be pointing to an INTERNAL attribution about her or an EXTERNAL attribution— need to know more, but something definitely is going on here! High consistency is the “engine” that drives an attribution—when consistency is high, it makes you more confident about your attribution/judgment).


            LOW consistency = No, you’ve never seen her act this way with him before. (Hmmmm…LOW consistency always causes sort of a mess in the attributional process—it means you’ve observed a “one-shot” deal. It’s hard to know what’s going on, unless it happens again. However, if you’re really motivated to make an attribution, you will, even with low consistency. The type of attribution will depend on the rest of the pattern (the other two types of information, Consensus and Distinctiveness).


Consensus information:


            How do other women act around this guy? Do they flirt with him too, or not? What’s the “consensus” here?

            HIGH Consensus = All the women at this party have been flirting with this guy, and/or, at past parties you’ve attended, a lot of women have been seen flirting with him. He’s just a babe magnet! (Hmmm….her flirting behavior with him doesn’t reveal much of anything about her, she’s just acting like almost every other woman who encounters him. This suggests an EXTERNAL attribution about her flirting behavior—it’s not her, it’s him that’s causing her to behave this way.)

LOW Consensus = No other women at this party have even talked to this guy; and/or, at past parties, all women avoid him. (Hmmm…this is really looking like an INTERNAL attribution—it’s something about her! She alone sees and appreciates his (hidden!) wonderful qualities!).


Distinctiveness Information:


            Is her behavior toward him distinctive? That is, does she behave toward him in a way that is different from how she behaves around other men in this type of situation? In other words, does she “flirt with all the guys” or is she typically shy and/or less out-going?


            HIGH Distinctiveness: At this party (or at past parties you have observed), she only flirts with him. She doesn’t flirt with other guys. In fact, her flirting with him is unusual, which is why you’re bothering to even think about all this! (Hmmm…this is suggesting an EXTERNAL attribution—it’s something about him that’s making her act this way—for example, he is just very appealing to her.)

LOW Distinctiveness: She flirts with all the guys at this party (and at past parties you have observed). She just can’t turn it off! It’s got nothing to do with this poor guy, he’s just her current “flirt object.”  (Hmmm…this is suggesting an INTERNAL attribution for her behavior with him—it’s all about her—she’s “a flirt”—her behavior has little or nothing to do with him.)


The Attribution:

So, you put all that information together, and the two most straightforward attributions that you can end up with are:


INTERNAL ATTRIBUTION: Her behavior is “all about her” (caused by her) (e.g., “She’s just a flirt—she’s not really interested in him!”)


            High Consistency    (She’s flirted with him every time she’s around him)

            Low Consensus   (No other women flirt with this guy)

            Low Distinctiveness   (She flirts with every guy she comes across)


OR:     EXTERNAL ATTRIBUTION: Her behavior is “all about him” (caused by him) (e.g., “She thinks he’s fabulous!”)


            High Consistency    (She’s flirted with him every time she’s around him)

            High Consensus  (All the women flirt with him)

            High Distinctiveness (She only flirts with him, not with other guys)