In this guest post, Dr Endre Tvinnereim of the Uni Research Rokkan Centre and Dr Kjersti Fløttum of the University of Bergen, Norway, shed new light on what people instinctively associate with the term ‘climate change’, and ask whether traditional survey questions may be unintentionally framing the debate on climate change.
Public opinion is crucial in determining what can be done to reduce emissions of greenhouse gases and adapt to climate change. At the same time, the topic of climate change touches on so many different themes (from the science to the societal implications) that what people associate with the phenomenon is highly variable and at times probably also fairly unsettled.
A research project conducted at the University of Bergen (just published in the journal Nature Climate Change) sought to shed some new light on the question of what people emphasise when they think about climate change. Using data from the 2013 online ‘Norwegian Citizen Panel’, we used ‘open-ended’ questions to collect the views of the public and use emerging text analysis tools to analyse the texts we receive – one from each respondents.
Open-ended questions are unusual on surveys, which tend to employ items with numerical scales, whereas the data we analysed were people’s written responses to the question ‘what comes to mind when you hear the words ‘climate change?’ This allowed people to identify themselves the aspects of climate change they thought were important, and permitted more nuanced answers than with traditional ‘closed’ questions.
The length of people’s answers ranged from one word to several sentences. People wrote these texts directly into their web browsers at home (pretty much all Norwegians have Internet access, making this kind of study easier and more robust). The topics expressed in the answers are diverse, ranging from melting ice at the poles, via more intense weather to effects on society and the role played by humans in causing climate change.
From the 2,000 answers that were collected, we found four overall themes, presented in order of declining prevalence:
- Weather and ice: an emphasis on the physical manifestations of climate change, notably through melting ice, more rain and storms (but not so far heat waves, this being Norway)
- Future consequences for humans, including concern for children and grandchildren
- Money and consumption, including both the negative effects of consumer society, the need to help poorer countries and hints about economic motives behind climate policies
- Attribution. What causes climate change: humans, nature, or a combination
The categories map fairly well onto the three IPCC working groups: physical science (topics 1 and 4); impacts (topic 2) and mitigation (at least some relation to topic 3). Topic 1 (weather/ice) accounted for almost half the responses; topic 4 (attribution) was the smallest at about 15 percent.
Older respondents had a tendency to emphasise weather and ice, whereas younger respondents were more likely to mention consequences for humans. We speculate that older respondents may have first learned about climate change in the 1980s when physical manifestations (such as melting ice sheets) were emphasised, whereas recent years have seen more discussion about the effects of climate change on human settlements, water availability, and food supply.
The words that were used the most were “extreme weather” (one word in Norwegian), “weather”, “warmer”, “natural catastrophes” and “human-made”.
There are several advantages to open-ended questions in survey research. Firstly, the open format enables the researcher to find out more about what people associate with a given subject, which is particularly useful when the subject is technically complex or multi-faceted. Closed questions may give precise numeric indicators of where people locate themselves on dimensions such as concern about climate change, but it is harder to gauge whether climate change to them is mostly about consequences, or about causes, or something else entirely (polar bears? electric vehicles? taxes?).
Open-ended questions allow us to access such thoughts more directly and to frame our analyses a different way: thus, the statements ‘climate change is mostly human-made’ and ‘climate change is not human-made’ could belong in the same category on our analysis, as both focus on causes (rather than, say, impacts). The way that researchers frame their analyses can have powerful effects on the wider discourse: could it be that ‘closed-ended’ questions have promoted the idea of a dichotomous split between ‘sceptics’ and ‘believers’ in human-causation of climate change
Second, allowing respondents to use their own words permits more nuances to make it through the survey matrix. notably, we found that among those most strongly wedded to the “attribution” topic, between one-quarter and one-third suggested there might be both human and natural causes of climate change. This held despite the fact that most of these respondents dismissed human causation of climate change in a closed question posed elsewhere in the study. It is thus possible that more varied and less polarised opinions may be uncovered when people are allowed to use their own words in survey research.