Monday 14 March 2016

Is there anything special about using character strengths?

Today we had a journal club discussion (presented by Hayley Jach) that substantially increased my skepticism about strength-based interventions. Here are a few quick thoughts on appropriate control groups in intervention studies, and a challenge to future strength-based interventionists!

Positive psychology interventions (PPIs) are intentional activities that aim to increase your happiness*. Of the different types of PPIs, strength-based interventions have been particularly well-researched. Broadly speaking, strengths can include positive personality traits as well as things that we're good at and enjoy doing (i.e., talents). The VIA classification (used in the paper I'll discuss) focuses on (moral) personality strengths, described by this taxonomy:

Importantly, although we all have all 24 strengths to varying degrees, apparently some strengths are more special than others. Signature strengths are the 3-7 strengths that best characterise us, and Seligman and Peterson (2004) have argued that using your signature strengths is invigorating, energising, and engaging. For example, if Kindness, Gratitude, and Love of Learning are your signature strengths, you'll be happier if you have regular opportunities to do things like donating to charities, writing thank-you notes, and attending talks.

In a seminal article reporting results from various PPIs, Seligman and colleagues (2005) found that participants who were instructed to use their strengths in a new way every day reported greater happiness and depression up to 6 months later, relative to a control group who were instructed to recall early memories.

Several studies** have since used this intervention and found various positive effects, but Proyer and colleagues (2015) decided to do something a bit different. Instead of just comparing the signature strengths intervention with a control group, they also introduced a second intervention group: using your "lesser strengths" (i.e., your bottom 5/24 character strengths).

They launched a large online intervention, beginning with 1,046 participants but ending up with 375*** who completed all follow-ups to 6 months, and were included in analyses. After taking the VIA character strengths survey, participants were given their intervention instructions. Participants in the two strengths intervention conditions were given the same instructions:

“We have selected five character strengths for you. Use one of these strengths in a new and different way every day for 1 week. You can apply the strength in a new environment or when interacting with a ‘new’ person. It is up to you how you want to apply these strengths. Try to apply these strengths, regardless of whether you feel like already using this strength frequently or not.”
The difference was that participants in the signature strengths (SS) condition were assigned their top 5 strengths, whereas participants in the lesser strengths (LS) condition were assigned their bottom 5 strengths (and neither group were not told which of these conditions they were in). Participants in the "placebo control" condition were instructed to write about their early memories.

The key finding was that participants in both strengths intervention conditions reported a tiny increase in happiness (but not decreased depression) for up to 3 months****, relative to participants in the control group.

From this, Proyer and colleagues concluded:

"working on the SS as well as working on the LS, is beneficial for increasing happiness."

And sure, it's plausible that maybe signature strengths aren't that special after all, and that using your lesser strengths can increase your happiness too (because it still involves acting in a more socially-desirable, or well-adjusted way).

But, the fact that we see identical effects in both intervention conditions suggests that something placebo-ish may be going on. So, before accepting Proyer and colleagues' conclusion, I think we should take another look at the intervention instructions: 

"Use one of these strengths in a new and different way every day for 1 week. You can apply the strength in a new environment or when interacting with a ‘new’ person."
Looking at these instructions, I am totally not convinced that these results provide any evidence for the effects of using strengths per se. Instead, the instructions seem to focus on getting people engaged, doing new things in life, and hanging out with new people (i.e., acting extraverted, which we also know is associated with increased positive affect). It seems to be more of a novelty and general behavioural activation intervention than a strengths intervention. Meanwhile, the "placebo control" participants are at home, writing about their early memories. It's just not comparable.

This is a challenge to the validity of all previous strength-based intervention studies that used similar novelty-based intervention instructions without an appropriate control group (including the seminal 2005 paper). 

A much more appropriate control condition would involve instructions that go something like this:

"Do something new and different every day for 1 week. For example, you can go to a new environment or interact with a ‘new’ person. It is up to you how you want to bring more novelty into your life."
So, before I can be convinced that using your strengths (signature or lesser) has any unique effect on happiness, over and above the effects of generally making an effort to be engaged in life, I want to see a replication of Proyer and colleagues, but with a control condition where participants have similar opportunities to be engaged with new and interesting things and people in life. I'd be very curious to see the results!


*more formally, "to cultivate positive feelings, behaviors, or cognitions" (Sin & Lyubomirsky, 2009, p. 468)

**of varying quality. do not get me started on various issues like power and representativeness of participants.

***huge huge attrition (typical of online studies). intention-to-treat analyses were not used.

****how plausible is it really is that a 1-week, one-off intervention could result in such long-lasting effects anyway? logically, it seems like sustained, long-term increases in wellbeing require similarly sustained, long-term changes in thoughts and behaviours, including habits and lifestyle choices. i'd be curious to know if this 1-week intervention motivated participants to get out there and be more engaged in life on an ongoing basis, beyond just the intervention week. that kind of information would make for a more compelling story! Proyer, R., Gander, F., Wellenzohn, S., & Ruch, W. (2015). Strengths-based positive psychology interventions: a randomized placebo-controlled online trial on long-term effects for a signature strengths- vs. a lesser strengths-intervention Frontiers in Psychology, 06 DOI: 10.3389/fpsyg.2015.00456

1 comment:

  1. Thank you very much for this post. As a life coach, I've found increasing pressure in the profession to use the VIA assessment to help clients "discover" their character strengths (the scare quotes indicate my suspicion that most people know very well what their main strengths are without answering a lot of general questions). I have many of the same reservations you express. I'm glad to read of the study you cite which, I agree with you, is not random enough. My thought is that in order to properly mimic the process by which people are encouraged to work on specific strengths, one method would be to randomize the characteristics and generate them using a computer program, or to have researchers pick blindly and assign five of them to each study subject, just as randomly. The notion that these five traits were "selected...for you" is probably psychologically important for the participants -- and, I suggest, would more closely parallel the usual suggestions than giving participants the option to select something new to work on every day. I believe that incorporating the notion that experts are giving subjects specific traits to work on is important to include in a study.

    By the way, I took the VIA assessment in 2014 (found it annoying and the results useless) and, in late 2017, took it again in case the advocates for VIA in the coaching world might be right. I still found it overly general, and the results useless. Then I realized I still had access to my 2014 results. When comparing the two, I found that of the 24 traits, only four remained in the same positions. Fully 25% shifted position significantly (that is, by three or more positions), the rest by one or two positions. To me, this suggests that there's nothing "core" or "essential" about these positions. Furthermore, some studies have demonstrated that, rather than focusing on the results of the VIA assessment, it is better for people to focus on those five or so traits that have been found to be most correlated with life satisfaction, such as gratitude, zest, love, and hope. In fact, other studies indicate that factors such as intelligence, gender, and neurosis correlate with certain traits more than others, and may skew the results of self-assessments. These, in addition to other more general critiques of self-assessments as objective measures, stand as strong arguments against the use of VIA and other such assessments as tools for raising the quality of an individual's life.

    I note that the VIA assessment and a perfunctory report of results is offered online for free, but that further reports and more in-depth discussions are only offered on a fee-paid basis.