by Michael D. Anestis, M.S.
Everybody has their own way of responding to negative emotions. Some of us work to change such feelings while others of us accept negative emotions non-judgmentally as a natural occurrence. Some of us brood over the meaning of what we are feeling and others of us shift our attention to other, more positive thoughts. Some of us turn to binge eating whereas others of us inflict bodily harm upon ourselves. A number of these responses are rather healthy and, in fact, are predictive of positive mental health outcomes, whereas others have been shown to make problems worse and to increase the likelihood of harmful outcomes.
On PBB, we have spent a fairly substantial amount of time talking about emotion regulation - the manner in which an individual attempts to alter or sustain an emotion. A large portion of that discussion has centered on behavioral methods for changing how we feel, such as binge eating, non-suicidal self-injury, and substance use. Additional articles, however, have touched on what is referred to as cognitive emotion regulation: attempts to change how we feel by thinking in particular ways (whether or not we are aware that our thoughts are serving that function at the time). For example, several earlier PBB articles discussed rumination, a thought style involving the tendency to brood over negative feelings.
In an article just published in Personality and Individual Differences, Kimberly Zlomke of the Baylor College of Medicine and Kathryn Hahn of Millsaps College (go Majors!) (2010) took a closer look at the different cognitive emotion regulation strategies used by men and women and how those strategies relate to a number of problematic outcomes, including stress, anxiety, and worry. To do this, the authors administered a number of questionnaires online to a sample of 1,080 undergraduates.
In total, the cognitive emotion regulation strategies most frequently used were positive reappraisal (reframing a negative event/emotion in a positive light), refocus on planning (shifting one's thoughts towards ways in which the situation can be resolved effectively), and putting into perspective (considering the negative event/emotion in the context and thereby not overstating its importance).
For men, the use of rumination (brooding over negative emotions) or catastrophizing (thinking about all of the bad things that could happen as a result of the event/emotion) predicted increased levels of worry and stress. On the other hand, the use of "refocus on planning" predicted lower levels of anxiety, stress, and worry.
For women, the use of self-blame (assigning fault to oneself), rumination, and catastrophizing was associated with higher levels of anxiety, stress, and worry. On the other hand, higher levels of acceptance and positive reappraisal predicted lower levels of worry.
What do the findings mean?
Based on this study, it appears that some cognitive emotion regulation strategies are harmful in both men and women. Specifically, rumination and catastrophizing were both associated with increased levels of worry in men and women. This is not surprising, as prior research has demonstrated that, while rumination is far more common in women than in men, it is just as harmful for men when they think in that manner (click here for our discussion of such research).
Perhaps more interestingly, the authors found that other strategies work differently in men versus women. For men, focusing attention on how to handle a negative event ("refocus on planning") was predictive of lower levels of worry, anxiety, and stress whereas for women, accepting a negative event/emotion and reframing it in a positive light were associated with decreased levels of worry. In other words, some of the cognitive emotion regulation strategies that work for men do not necessarily work for women and vice versa.
No doubt, findings like these are interesting, but they need to be considered with care and subject to a reasonable amount of skepticism. Almost any time you compare two large groups of people on a large number of variables, the groups will differ from one another statistically on some of those variables. Some of those differences are due purely to chance whereas some of them are true differences. Even some of the true differences are questionable, however, because statistical significance in a large sample does not necessarily reflect a large difference. All statistical significance means is that the results are not likely due to chance (more specifically, that there is a less than 5% chance of finding the results purely by chance), not that the differences are big.
All of that being said, if you are going to make comparisons like this, which is a worthy venture, it is important to have guesses made ahead of time as to how the results will end up and evidence from prior studies that support your rationale for making those guesses. In other words, we need to have a sense of why something happened and be able to use that justification to predict future results rather than simply looking at something that happened and attempting to rationalize it. Given the lack of research in this area, Zlomke and Hahn should not be faulted for taking a more exploratory approach, but we would need to see these results replicated before we could be fully confident that they reflect a robust series of relationships that generalize across samples.
As it turns out, I have collected multiple data sets on undergraduate samples using the same primary measure - the Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski et al., 2001). In an effort to see whether or not the findings from Zlomke and Hahn (2010) were due purely to chance, I took a quick look at one of those data sets a moment ago and found the following results (the numbers represent the average score for each group on each CERQ subscale):
|Refocus on Planning||Positive Refocus||Positive Reappraisal|
|Put into Perspective||Catastrophize||Blame Others|
Taking it a step further, I looked to see whether any of the differences between men and women on these scores were statistically significant. None were, although two approached significance: self-blame (p < .078) and rumination (p < .06).
My next step was to see which of these subscales were significantly related to depression and anxiety for men and women. For men, higher levels of refocusing on planning and positive reappraisal were associated with lower levels of depression. Also for men, higher levels of rumination predicted higher levels of anxiety. No other subscales were related to depression or anxiety.
For women, higher levels of self-blame, rumination, catastrophizing, and blaming others were associated with higher levels of depression and anxiety. Surprisingly, higher levels of acceptance were associated with higher levels of depression. Also for women, higher levels of positive refocus, positive reappraisal, and putting into perspective were associated with lower levels of depression and anxiety and higher levels of refocusing on planning were associated with lower levels of anxiety.
Now, to be fair, my sample size was only 200 people and approximately two-thirds of my sample was female, so finding significant differences between the groups is a bit difficult. At the same time 200 people is not tiny, so these results should not be dismissed out of hand. As you can see from my results, there is reason to be somewhat skeptical of the findings of Zlomke and Hahn (2010).
In my sample, there were no significant differences between sexes on the use of particular cognitive emotion regulation strategies, although the small sample size may have obscured some differences and, indeed, self-blame and rumination were close to the threshold for significance (p < .05). Additionally, the manner in which different subscales are associated with problematic outcomes was very different than what was reported in the Zlomke and Hahn (2010) paper. Now, to be fair here too, I used related but different outcome measures (the Beck Depression Inventory - II and the Beck Anxiety Inventory) than Zlomke and Hahn (2010) did, but the differences are still hard to overlook. Not only were the results not identical, some completely contradicted one another (e.g., in my sample, acceptance was associated with more depression whereas in their sample it was associated with lower levels of worry).
So what should you take away from all of this? First of all, it is interesting to consider the ways in which we manage our emotions through styles of thinking and it is more than reasonable to see if men and women differ in their strategies and if the results of those strategies differ for men and women. At the same time, if we simply run a large number of exploratory analyses in a large sample, the degree to which we can be confident that our results mean anything is very limited. When I ran a series of similar analyses with my own data, based upon a smaller but demographically similar sample, my results were very different. How do we know who is right? The answer is replication, replication, replication. There is interesting information to learn here, but to learn it we need to see how these things shake out over time when the same tests are run several times by different people independent of one another. What we think we learn from one data set needs to be tested in a new one in order to see if our new theories can predict results we have not yet seen or if those theories simply reflect an interesting but misguided attempt to explain something that already happened.
If you would like to learn more about rumination and worry, we recommend the following items, which are available through our online store of scientifically-based psychological resources:
- Women Who Think Too Much: How to Break Free of Overthinking and Reclaim Your Life
by Susan Nolen-Hoeksema
- The Worry Cure: Seven Steps to Stop Worry from Stopping You by Robert Leahy