4 ms·

You bring up sample size here which doesn't really matter because the effect is statistically significant. Your point about the populations being unrepresentati

• pak 10y ago
  The problem is you're arguing against a totally different position than where Geekette and myself are coming from. Let's go back to the first lines of the OP: "What are the ingredients of a good relationship? Trust? Communication? Compromise? How about a sense of smell?" ... followed by the statement by Dalton that "Smell is important in social bonding." That last statement there is an extremely broad, confident statement (consider how complicated social bonding is), and it's explained by Dalton with this "emotional contagion" idea. Two studies that show statistically significant effects for a total of ~60 people smelling odors from two different conditions (fear and disgust) is not nearly enough evidence to back up an assertion as broad as "smell is important in social bonding," nor that a substantial range of emotions spread via odors (contagion). At best, these two studies are suggestive, at worst they are so flawed as to not lend any evidentiary value. "Statistical significance" is not a substitute for critical thought on how much evidence stands for and against the hypothesis (meta-analysis). If I set up two experiments that showed that gravity doesn't exist, p=0.04 for both, would you set aside the totality of the evidence saying it does? Perhaps you would criticize my methods, or the link from my results to my conclusions, no? You set up a strawman by saying that I'm asking the impossible—that large-scale behavioral cohorts must be assembled for preliminary studies. I'm not. I'm simply expecting some honesty in how people report the conclusions of preliminary studies. >Did you even read either abstract? This is controlled for. There is a control group inhaling normal sweat You are incorrect, probably because you did not read past the abstract. Perhaps you should be more careful about telling people how science works in the real world or whatever, because real scientific discussion involves reading methods sections. In the APS paper contrasting fear vs. disgust vs. control, the control condition contained unused compresses, i.e., no sweat at all. From the horse's mouth: "unused absorbent compresses [...] in our view constitute optimal control stimuli because other nonemotional bodily secretions (e.g., sweat from playing sports) can potentially contain other chemosignals." So... I am supposed to believe that when recipients smelled sweat from the donors watching disgusting videos, compared to no sweat, feeling disgusted at the sweaty odor is some kind of "emotional connection" to how disgusted the donor felt while sweating? Give me a break—this is a sham negative control. The huge confounder, as Geekette rightly points out, is that many people are normally disgusted by any sweaty odor.
  • ASpring 10y ago
    Yes, let's go the OP that I addressed. Nowhere have I made claims supporting the original articles conclusions. Only claims supporting emotional contagion that was dismissed out of hand by Geekette. >I think the concept of “emotional contagion” is a stretch. I simply do not believe that empirically[1], people who smell sweat excreted by frightened people will also feel fear. Is this a reasonable statement without doing any prior research? This is a rejection without reason given prior evidence to the contrary. My response was simply that it's absurd to reject an entire idea without doing any prior research. I'm sure you'd agree with that so I have to imagine we're just engaging in mental grad student masturbation that we've all been trained to do. But let's address your points. > Two studies that show statistically significant effects for a total of ~60 people smelling odors from two different conditions (fear and disgust) is not nearly enough evidence to back up an assertion as broad as "smell is important in social bonding," I, nor the studies I linked, claimed this. >"Statistical significance" is not a substitute for critical thought on how much evidence stands for and against the hypothesis. Evidence against has yet to be presented. >If I set up two experiments that showed that gravity doesn't exist, p=0.04 for both, would you set aside the totality of the evidence saying it does? This is an absurd hypothetical. Our prior for knowledge about emotional contagion via smell is uninformed. We see two studies supporting the idea, can we conclude the idea is true? Of course not, but it's even more absurd to claim the exact opposite as Geekette did. >You set up a strawman by saying that I'm asking the impossible—that large-scale behavioral cohorts must be assembled for preliminary studies. I'm not. I'm simply expecting some honesty in how people report the conclusions of preliminary studies. We've had a bit of misunderstanding here. You argue against the original article and presume I'm defending it. I'm not, it's mediocre reporting that widely extrapolates these studies. All I suggested was that dismissing something out of hand is absurd without doing research. >You are incorrect, probably because you did not read past the abstract. Sorry I confused the controls in the studies. Check the other which used a sport sweat control. As fun as this is I'm not going to continue past here.