My empirical work has mostly focused on the emotion of envy. I have a lot of data that I need to publish. I've been frozen lately by a sort of methodological crisis and metamorphosis. I started thinking about a number of validity issues after working on the BBS paper, and I couldn't move forward with some of my research until I sorted out whether it was valid or useful. (I'm as hard on my own methods as I am on everyone else's.) I'm trying to integrate much of it now, and I thought I'd share some interesting things I've learned in trying to uncover some causes and predictors of envy.
So far recalled experience has been much more useful than attempts to elicit envy with canned inductions.
A recalled experience is what it sounds like – ask the participant to recall a time when they experienced envy toward someone and have them write briefly about it.
To elicit a new envy experience during a study, I've had people read various narrative passages or articles. I stopped using student samples a few years ago, but I have lots of student data from before, and with ASU students I had crafted fake student newspaper articles about a particularly affluent student (gender matched), for example.
More recently in looking at envy and anti-Semitism, I've used passages that report some of the ways in which Jews excel, from higher mean IQ scores, to household incomes and Nobel Prize take rates.
The basic structure with these inductions is that you have the participant read the article and then you're looking for envy downstream, by direct measures (e.g. "How envious do feel right now?" sandwiched within other emotion items like anger, happiness, etc.) or oblique measures (mostly the kinds of things discussed by van de Ven et al.)
I've never seen a lot of envy with the canned inductions. Maybe 10 to 20 percent of the sample will report any level of envy in that context.
However, with the recalled experience method, most participants will have something to say about a past envy episode. This method requires trained coders to extract whatever dimensions you're interested in (or you could automate it with text analysis software), but it gives you much richer, and I think, more valid, information.
In fact, I wouldn't assume that "envy" is commensurable across these methods. Meaning, that what we're calling envy in one method may not be the same thing that we're measuring in the other. The people who report envy after reading about an ASU student, or Jews, might not be the same people, in a sense, as those reporting vivid envy episodes in a recall exercise. That is, the emotional experience might be substantially different, and the personality predictors different as well.
One of the most important lessons Lani Shiota taught me is the benefit – and sometimes the absolute necessity – of measuring emotions without referring to those emotions by name. ASU gives you excellent evolutionary psychology training. Well, I criticize the evo psych framework sometimes, but it makes us better social psychologists than we would be without it – even if we don't fully agree with the framework or some of the hypotheses generated by it. The evo psych perspective on emotions looks at their function, or their fitness value in the ancestral environment, and so when we want to carve nature at its joints, we don't want too many modern words and abstractions in the way.
Now, I actually used the term envy in asking people to recall an episode. I didn't say "Think of a time when someone else had more than you did, or had something you badly wanted, and it made you feel bad." Rather, I asked explicitly for an envy experience. However, the open-ended nature of the recall method gets you much closer to carving nature at its joints. You see a broad range of elicitors and situations. You see the role of gender in ways that would be hard to pick up with canned inductions. You're not as restrained by your priors. (Yes, there's a potential trade-off with effect hunting, cherry picking, multiple comparisons, etc, but I think those can be managed. Exploratory research is extremely promising and should be more common.)
I never thought about this, predicted it, or had any hypotheses about it. It simply never occurred to me. I'm the oldest of three, and the only boy. We're evenly spaced about six years apart. I have no experience with sibling rivalry – I've never been in competition with my (much younger) sisters, and have never resented anything they've done or achieved. I'm a low or zero envy person in general – perhaps another reason to prefer the recall method and be open-minded.
Sibling rivalry is real. I was struck by some of the participants' accounts.
The most striking pattern is that some women really hate their sisters. It was heartbreaking to read – there's real pain out there, people suffering greatly from the achievements or life status of their siblings (especially their sisters.)
One question in the recall exercise asked people what they wished had happened to the person back then, or what they wished they could have done to the person. One woman reported that she wanted to "beat her bloody", referring to her sister. The circumstance that evoked the envy was that her sister had a "good husband" and a nice house. Her sister hadn't actually done anything to her as far she reported – just having a good husband and being affluent (perhaps because of her husband.) Well, that might be biased on my part. She might maintain that her sister had done something to her – it just may not look like a thing that was done to her to a man like me.
This probably highlights some of the pressures and expectations women in our culture place on themselves and/or we place on them. Some women are defining themselves through the men they marry. Perhaps that's too strong – maybe it's more that finding a good husband is a major box checked for some women, and perhaps more time-critical, than the converse for men. I think this would be much less true with academic women, but I suspect it's more intuitively accessible to them than it is to men. Women know women.
This is actually a good example of the value of multidimensional diversity in any social science or branch of research psychology. We'll be limited in our ability to understand major swaths of human motivation if we don't experience them ourselves or come from the requisite culture. The BBS paper focuses on political diversity, but I increasingly think other dimensions are just as important – we need a lot more black people, Latinos, Natives, people from rural communities, people who have served in the military, people from non-Western and non-affluent cultures, and probably more masculine men. (I don't think "men" and "women" are necessarily sufficient categories. Semi-relatedly, GLBT are not underrepresented in social psychology – they're overrepresented (the percentage in our field is greater than the percentage in the general population), which is why I didn't mention them. Maybe we still need more. I'll have to think about this.)
There will always be selection biases. Human populations vary on all sorts of dimensions, including interest in things like social psychology or football. For example, maybe ex-military types will be less interested in being social psychologists (though I think we might be surprised – understanding human behavior is somewhat central to military life, and the nature of leadership and performance are constant areas of focus.) I think we could do a lot to make our field more inviting to diverse groups. The culture of social psychology is far too narrow and specific. I don't think a good social science can be so culturally specific (in our case, white/Asian urban liberals, overwhelmingly) – I think the very nature of social science requires breadth and diversity on multiple dimensions. The out-of-nowhere sibling rivalry discovery made me think about this again, how my personality, my background, my birth order, and my gender had made me blind to a major envy elicitor...
José L. Duarte
Social Psychology, Scientific Validity, and Research Methods.