We concluded last time by raising the issue of how self-selection may muddy the waters in happiness/well-being-related research.
Self-selection in this context refers to a scenario in which researchers seek out participants for well-being research and attract responses from individuals who are interested in cultivating personal well-being and boosting their happiness.
I refer to this as being potentially problematic not because having people who are interested in well-being knowingly taking part in well-being studies somehow invalidates whatever findings are reported; instead, the issue relates to generalisability. For the most part, in psychology, research findings are regarded as being most valuable when they can be applied to as wide a population as possible. So, if a group of researchers conducts a particular study with participants drawn from the general population and reports how they went about it in a step-by-step manner (in line with the scientific method), then another researcher in another city or country should be able to replicate that procedure and, ideally, report similar findings.
The catch with self-selection is that it can limit the potential for generalisability. I referred in the previous paragraph to the ‘general population’, and did not do so casually. Unless researchers are targeting a specific subset of the population (i.e., those with a certain medical history or of a particular ethnicity), they will tend to seek out representative samples of a given population, paying heed to demographic factors such as age, gender, education, and income. Doing so allows researchers to suggest that their findings may be relevant to wider society.
The specific pitfall that happiness/well-being research can run into is that researchers may think their participant population is representative (and it may tick the necessary demographic boxes), but what if they have attracted mainly individuals who are interested in the research topic in a way that may not reflect the preoccupations of the general population?
At that point, it is not that the research has no value, but instead, it may well be that findings arising from such studies may be generalisable only to those of us who prioritise cultivating our individual well-being, as opposed to the wider society.
That is all well and good as far as it goes, but it raises another potential research limitation that can be difficult to measure. Namely, to what extent can you be sure that self-selecting individuals convinced by the merits of well-being interventions and eager to benefit from participation are or are not influencing results? Specifically, if individuals are somehow primed to benefit, to what extent might their positive results owe more to their eagerness than to the objective merit of the intervention itself?
It was precisely this type of question that Sonja Lyubomirsky, Rene Dickerhoof, Julia K. Boehm, and Kennon M. Sheldon sought to answer in a research article published in the journal Emotion in 2011.
With a view towards measuring how individual motivation can impact on results, the researchers advertised the study in two different ways, making it look as if each was a distinct project. It was advertised separately as being either a ‘happiness intervention’ or as a study involving ‘cognitive exercises’. The researchers explained that this led to participants effectively self-selecting into either ‘high motivation’ or ‘low motivation’ groups, i.e., those who signed up for cognitive exercises were not as motivated to become happier as those who knowingly signed up for a happiness intervention.
The 355 United States-based undergraduate university students were then told – irrespective of which advert they responded to – that the purpose of the study was to increase well-being. They were then randomly assigned to one of three groups – optimism, gratitude, or control.
Participants in the optimism condition were asked to spend 15 minutes every week for a period of eight weeks writing about a future in which they were living their ideal life. The instructions offered here were very much along the lines proposed by Laura King’s Best Possible Self approach (as outlined in an earlier post).
Those assigned to the gratitude condition were asked to spend the same amount of time thinking about moments in life when they were grateful for something that another person did for them and to write a letter to that person, but to not send it.
Participants in the control condition, whose primary function was to provide a comparison point for changes over time, were asked to spend 15 minutes per week writing lists of what they had done over the previous seven days. With a view towards convincing these participants that they were in an active condition, researchers told them that this activity could impact positively on well-being by improving organisational skills.
Well-being variables (e.g., happiness and life satisfaction) were measured on three occasions – at the start of the process, at the end of the eight weeks, and six months later.
They reported that self-selection appeared to influence the results, in that the greatest increases over time were reported by those assigned to the optimism and gratitude conditions and who had initially signed up for what they thought was a happiness intervention. The logic here is that they were more motivated to become happier than those who thought they were signing up for the cognitive exercises and that this seemed to be reflected in the outcome.
However, they stressed that there was no difference over time detected among participants assigned to the control condition. This is important, because it counters the idea that priming alone accounted for the positive results in the active condition, i.e., if being motivated and therefore primed to benefit was the sole driver of the outcomes, then highly motivated participants in the control condition should also have fared measurably better than their counterparts, but this was not the case. That, in turn, suggests that well-being and happiness-boosting interventions can have objective benefits, as opposed to those positive results amounting to a placebo effect.
Reflecting upon their findings, the authors offered the conclusion that while well-being interventions are not placebos, they will tend to be most effective when people know what they are about, are motivated to benefit, and commit to whatever activity is being asked of them.
Dr. Mark Barry
Mark Barry was awarded a PhD by University College Cork in 2015 for his research into adolescent well-being. He has lectured psychology at UCC since 2013, and is also a freelance writer.