Home / Topics / Research Design / Measurement / Measurement Development / Onion & Garlic Symptoms
Eric A. Youngstrom

Onion & Garlic Symptoms

Posted on November 15, 2007

Eric A. Youngstrom (bio) considers improving bipolar disorder measures to address discrepancies among informants.


There’s no one best way to handle getting information from multiple sources, and I feel, honestly, like I’m setting people up when I’m teaching a class and saying, “You should gather information from multiple informants,” because it’s never going to agree. It will agree slightly more often than chance.

The correlations are a .2, .3; so occasionally you get lucky and the stuff seems to line up. But usually it doesn’t, and the more that you add to the mix the more complicated the picture is.

The flipside of that is also the richer an understanding that you have, and we’re starting to learn that some of the discrepancies — we’re starting to learn where some of them come from. Concretely in the RO1, we’ve given the teenager and their parent exactly the same questionnaire to fill out, and we try to predict who is going to have bipolar and who isn’t going to have bipolar disorder. Mom knows best. You can ask exactly the same questions and the parent is doing better.

And I think talking with the families and doing some follow-up digging, there are two things that are really driving that process. One is what Kay Jamison and others have called the Seduction of Mania and Hypomania.

One of the features of this is that you lose insight into your own behavior. You don’t feel ill; you feel good. You feel better than good. And your behavior starts to bother other people before it starts to bother yourself.

Denny Cantwell and Gaye Carlson have talked about this as onion versus garlic symptoms. Onion symptoms would bother me if I was having them, and garlic symptoms bug everyone around me when I’m having them. And mania is chockfull of garlic symptoms.

So going back to that questionnaire, there’s an item that asks the parent, “Does your child get more irritable than usual, particularly with changes in their energy?” And mom will say, “Yeah.” And you ask exactly the same question to the teenager and the teenager says, “No, what a stupid question.”

And so they’re circling a no on the questionnaire, but looking at the discrepancy now as a clinician, I can ask the question differently. I can look at that response and say, “No, it looks kind of like an irritable response.” “Well, that’s your issue; that’s not mine.” I’m like, “Okay.”

So clinically, the questionnaire can’t change the way that it’s written, but as a clinician, I can ask questions — like instead of asking it, “Do you feel more irritable?” Say, “Do you notice people getting on your back more than usual? Do you feel like you’re surrounded by idiots more than usual?” And that’s really putting myself in the person’s shoes and approaching it.

What would be neat would be to write a second generation questionnaire which asks the questions instead of asking, “Do you feel more irritable, Do you feel surrounded by idiots?” There’s a group in Sweden that started to do that with antisocial behavior, where they're asking about psychopathic traits in a way that sounds cool if you’re a psychopath.

It doesn’t ask, “Do you lie a lot?” It asks, “Can you get everyone else to believe pretty much whatever you want anytime you want?” And so I think that that sort of — and again, that’s a neat example, to me, of what this sort of science practice mixture could create is taking that clinical insight about how to ask the questions better and develop a better questionnaire that we could use to inexpensively get closer to the truth.

 

« Back to Article