quote:

ORIGINAL: Anaxagoras
I don't see quite how the secondary questionnaires could assess the truth of the answers on the initial questionnaire relating to the emotional states of their partners, unless their own emotional state was similar to that which they were asserting about the partner but that would not negate truth value.

quote:

If the above is true, that empathy was the sole (rather than predominant) element being analysed in the study, then I would appreciate a quote from the study to verify that. As I understand it, they would have been unable to measure statistically how for example women expressed more satisfaction over the attempts of their partners to be empathetic, while their satisfaction levels were less strongly linked with actual empathetic accuracy, if the study did not factor in the personal feelings of each partner beyond that of empathy.

quote:

I can see that the article headline might be misleading but not that it makes no sense, for the same reasons that I explained previously. It uses words like "happy" in an unqualified fashion when it should relate to satisfaction within relationships.

quote:

My own view is that it is unreasonable to expect the same standards of accuracy as one would have in a academic journal but that should not preclude the body of the article being a proper representation of the truth, and in that respect I think the article does quite well.

quote:

ORIGINAL: shallowdeep
There might be some lingering misunderstanding. Both partners reviewed the same segments of the conversation. For each each segment, they both filled out two questionnaires: one representing what their they thought their partner was feeling during the recorded segment and the other recalling what their own feelings had been when the segment was recorded. The authors then took what the woman thought the man's feelings had been during the segment and compared it to what the man actually reported he was feeling at the time, i.e. the "truth." Similarly, what the man thought the woman was feeling was compared against the emotions the woman actually reported feeling during the same segment. This enabled a test of empathic accuracy and was the only use of the questionnaires on emotions. Note that there is no data presented about what the recorded emotions actually were; what participants were feeling during the segments is irrelevant to the study, only the participants' accuracy in judging their partners' emotions is being examined. Does that make more sense?

quote:


Use of the electronic device was limited to selecting particularly emotionally-charged moments (i.e. the HAMs) from a longer original conversation for further analysis with the questionnaires. That was the extent of its application; no data or analysis of data from the device are reported.

quote:

quote:

If the above is true, that empathy was the sole (rather than predominant) element being analysed in the study, then I would appreciate a quote from the study to verify that. As I understand it, they would have been unable to measure statistically how for example women expressed more satisfaction over the attempts of their partners to be empathetic, while their satisfaction levels were less strongly linked with actual empathetic accuracy, if the study did not factor in the personal feelings of each partner beyond that of empathy.

As stated in the abstract's first line, the study was only intended to examine links between three categories: existing relationship satisfaction, empathic accuracy, and perceived empathic effort. The full matrix of correlation coefficients from the study data is available in Table 1. Those are the only links for which data is presented. Note the lack of anything about the emotions themselves.

quote:

ORIGINAL: shallowdeep
One thing that I think may be confusing you, and perhaps the Telegraph, is the the meaning and use of the "relationship satisfaction" measure. That measure was not related to the questionnaire about emotions or, indeed – providing I'm interpreting it correctly – to the experimental, emotionally charged conversation at all. The level of relationship satisfaction was determined using the Locke–Wallace Marital Adjustment Test apparently commonly used for the purpose, probably with a form very similar to this one. Given that the test was chosen, in part, for test-retest stability and that the methodology does not indicate it was given with each HAM segment (as it does with the other measures) the implication seems to be that this survey was administered just once, to determine a participant's existing, long-term satisfaction with their relationship. This long-term measure of relationship satisfaction does not seem intended to reflect any short-term effects of the experimental interaction. Instead, based on the idea that the experimental snapshots are a representation of the ongoing empathic accuracy and perceptions of empathic effort present in the relationships, the study looked for correlations between those factors and the participants' reported long-term satisfaction with their relationships.

I could be wrong, of course. Perhaps the study actually was investigating how, after years of marriage, 30 seconds would make a couple from Bryn Mawr suddenly reconsider who they would marry if they had their life to live over again…

quote:

To test the relative contribution of empathic accuracy and perceived empathic effort to self and partner’s satisfaction, we used the Actor–Partner Interdependence Model (APIM; Kashy & Kenny, 2000), which analyzes data from both partners simultaneously and thereby helps to distinguish between the individual and dyadic influences of each empathy variable on each partner’s satisfaction. For example, a woman’s empathic accuracy may lead her to feel more satisfied in the relationship (an individual effect)
and/or it may lead her husband to feel understood and more satisfied in the relationship (a dyadic or partner effect).

quote:

ORIGINAL: shallowdeep

quote:

ORIGINAL: Anaxagoras
I can see that the article headline might be misleading but not that it makes no sense, for the same reasons that I explained previously. It uses words like "happy" in an unqualified fashion when it should relate to satisfaction within relationships.

It is actually the "when" in the headline that is more problematic than the "happy." To support the claim, it would be necessary to show that there was some significant correlation between a woman's happiness (or, to fix that bit, her satisfaction with her relationship) and her partner being upset. No correlation presented supports that. In fact, since negative emotions themselves are not a category examined for correlations in the study, it's not possible to even attempt to draw that conclusion from the data.

quote:

quote:

My own view is that it is unreasonable to expect the same standards of accuracy as one would have in a academic journal but that should not preclude the body of the article being a proper representation of the truth, and in that respect I think the article does quite well.

While I certainly don't hold popular publications to the same standard of depth or precision as an academic journal, I don't think it's unreasonable to expect general accuracy in the information that is published. If a reporter lacks the time or background to understand academic publications, they really shouldn't be reporting on them. If the Telegraph just used a sensational headline, I might roll my eyes, but I wouldn't take any particular issue with it. In this case, however, I think it's worse than that. The novel additions to the press release serve to mischaracterize the research to the point that a reader will likely take away something very different from what the study supported. I don't see value in that type of science reporting. The paper, or at least its readers, would have been better served simply republishing the press release than having Mr. Orr attempt to edit it.