Mindfulness and sustainable behavior: how to find a correlation where none might exist

Since recently I discovered that there is a whole field of what is called ecopsychology. A couple days ago I started reading a paper by Amel, Manning, Scott and Koger (probably more about that later) and I wondered whether those four had previous papers as well on the subject. Apparently they had. Members of this group produced a bunch of papers with (an)other member(s). One of the those papers had the intriguing title “Mindfulness and Sustainable Behavior: Pondering Attention and Awareness as Means for Increasing Green Behavior” by Amel et al (2009). It was quite an intriguing read, hence this post.

The authors started from the observation that our rushed lifestyle separates us from nature and this let us fall back on automatic behaviors which are not necessarily sustainable. They investigated two aspects of mindfulness: acting with awareness (paying attention) and being in the here-and-now (observing sensations). The authors hypothesized that paying attention is necessary for making sustainable choices and their goal is to break through this automatic behavior with mindfulness, so people could adopt a more sustainable behavior.

The conclusion of the paper was that, indeed, “acting with awareness is significantly positively correlated with self-reported sustainable behavior”. They arrived at this conclusion by means of a survey. Participants completed two questionaires. One to test their level of mindfulness (they investigated two aspects: “acting with awareness” and “observing sensations”) and the other to test their level of how “green” they were. Greenness was measured on a scale of 7: 0 being “not green” (meaning: never choose the most sustainable option available if it’s more costly in terms of time, money, convenience, or personal preference) to 7 being “dark green” (meaning: always choose the most sustainable option available, even if it’s more costly in terms of time, money, convenience, or personal preference).

According to the authors, there was no correlation between “observing sensations” and the “self-reported Green Scale ratings”, but when they put “acting with awareness” against the “self-reported Green Scale ratings” in a graph, this was the result:

I was not particularly impressed. Those who score low on the Green Scale, score all over the place on the mindfulness scale. Also, although a diagonal area is recognizable, most of the area of the graph is filled with values, except for the top left area (low mindfulness/high greenness) that is empty. The authors also noticed this and took it as an indication that their hypothesis could be right:

Our hypotheses were partially supported in that the acting with awareness subscale of mindfulness was significantly positively correlated with sustainable behavior as measured by the Green Scale. The actual pattern of the data is particularly interesting as there are no outlying data points in the low mindfulness/high self-reported green ratings area. This suggests that mindlessness is incompatible with sustainable behavior.

I am not so sure about that last sentence. I see no reason why low mindfulness can’t coincide with sustainable behavior. That would mean people who theoretically know what actions are sustainable, but live their life on automatic behavior. When I look at my own history, I knew from my education what was sustainable, yet looking at the examples of what is mindful, I probably would not score very high on that scale. I guess that my data point would be somewhere in that now empty area. I can’t imagine that I would be the exception. So I began to wonder why that area was so empty.

That became a bit clearer when looking at the methodology. The survey was conducted at an expo and visitors of that expo were invitated to take the survey. The expo was a “Midwestern sustainability expo”. They also mentioned its name: the Living Green Expo (LGE) of 2006.

So, if I understand that well, they were searching for the relationship between sustainable behavior and mindfulness on an expo where the large majority of the fair visitors are interested in sustainability? The chance that there are people who are not “green” will be very small. That means that only part of the relationship will be established. This means that they are mining results in the top half of the graph and much less data points at the bottom.

Remarkable, there were participants who were not interested in sustainability on that sustainability expo. Look at the mindfulness data points of 0 and 1. Even more, three of those data points were from participants who scored very high on the mindfulness test, but very low on the greenness test. They can be found below right on the graph. The authors noticed those also and found an explanation: those participants were family members of visitors of the expo and didn’t visit the expo because they were interested in sustainability.

The authors considered them “outliers” and tried to find all kind of reasons why they shouldn’t be on the graph:

Our current definition of mindfulness is restricted to a limited subset of the cognitive realm, specifically attention and awareness. In addition to attentional mindfulness, Wallace and Shapiro (2006) described conative mindfulness as “reflection on meaningful and wholesome desires, recognizing unwholesome desires that will lead to suffering both for oneself and others” (p. 694). Ecopsychologists have pointed out (e.g., Sewell, 1995) that a deep connection with the natural world is likely to begin a love affair with the earth that will make sustainable choices seem more critical and thus hard to ignore. Indeed, it is possible that our outliers are mindful in a cognitive but not an ecologically connected sense.

and

This suggests that mindfulness may be necessary but not sufficient in predicting green behavior. We took a closer look at the surveys belonging to these outliers and found some systematic patterns among their demographics. The outliers, who were all brought to the LGE by a family member, rather than attending because they were interested in the event, rated themselves low on a stages-of-readiness scale that had been included in the survey (response options ranged from “I have not heard about environmental problems” to I “have heard of environmental problems and have completely changed as a result”). These particular respondents either did not believe that there are environmental crises or they believed it but were not ready to change.

I don’t know whether these were just outliers or should these be considered an indication that something is not really right with their hypothesis. I was even a bit surprised that there are so many data points in that high mindfulness/low greenness area, knowing that the survey is conducted in a place mainly visited by people who are more likely to be interested in sustainability.

So, are these data points outliers because this combination is uncommon in the general public or because it was measured in a group that has only few participants with low scores on the Green Scale? It gives me the impression that the authors expected a correlation and in that case, of course, those three data points could be considered outliers. I wonder how much the correlation would be if the survey was taken outside the expo, with a better mix of people having varying interest in sustainability?

Back to the issue of this empty space of low mindfulness/high greenness. Luckily, the “caveats” section was well documented, kudos to the authors for that. In that caveat section I found that although there were 14,000 visitor of the expo in 2006, the organizers had a problem finding volunteers to take the survey. Many fair visitors who were invited to take the survey were “too rushed and did not have time”… This is an interesting observation because these expo visitor are more likely to score low on the mindfulness, yet being interested in sustainability. Therefor would likely populate that now empty area that was given as part of the explanation why that mindfulness correlates positively with sustainable behavior.

They were not only mining for people who score better on the Green Scale, but apparently also those who score high on the mindfulness scale. I wonder how much the correlation would be if these were included.

Concluding, how do you find a guaranteed correlation between mindfulness and sustainability, without there having to be an actual correlation in the population? The paper shows that it is rather easy to do:

1. Conduct a survey in a place where people who are interested in sustainability come together → this will leave you with very few participants who will score low on the sustainability scale
2. Ignore those people who show the hallmarks of mindlessness → this will leave you with very few participants who will score low on the mindfulness scale
3. Now look at the data. If you don’t find data points in the low mindfulness/high greenness area or in the high mindfulness/low greenness area, then, hallelujah, this is of course exactly what you expected!
4. If you nonetheless find data points there (there will not be many anyway, because of your methodology), then don’t worry, just consider these to be outliers…

Congratulations! Now you have found a significant positive correlation between mindfulness and sustainable behavior!

Advertisement