About the five pragmatic principles of a scientific qualitative approach
Dominique Desjeux, anthropologist, Professor Emeritus at the Sorbonne, University of Paris, CEPED, international contract researcher
Should the use of “chloroquine” or its derivative be made widespread, as recommended by Professor Raoult, or not?
A qualitative anthropologist’s point of view.
In the face of the urgency and human distress caused by the death of our loved ones by VIDOC-19, one might be tempted to expand the use of this drug. Why not, and that is what the government has just decided. This does not prevent us from thinking scientifically in order to make better human decisions about the question posed by a 24-person trial, that is, a qualitative survey.
As an anthropologist I am used to working on 10 to 70 people per survey. Clients regularly ask me if these are serious investigations. This is the same remark that is made with respect to Professor Raoult. The permanent doubt which exists relating to qualitative surveys has allowed me to reflect on the scientific criteria and the generalization of qualitative approaches. This reflection is partly, but not entirely, transposable to the question of hydroxychloroquine, provided, as always, that other approaches complement or discuss this reflection methodologically.
In practice, qualitative approaches are based on five scientific principles.
1 – The rule of “generalization of significance”, as opposed to a “statistically representative generalization” linked to a series of correlations.
2 – Induction, which means a practice of exploring the problem with very few hypotheses associated with a deep description of the phenomenon.
3 – The rule of ambivalence, which shows that every reality has a positive and a negative side. It is at the heart of political decisions, which have the role of adjudicating between these different sides.
4 – The fourth rule is the comprehensive approach. It tries to avoid value judgements and denunciations. It is a critical approach that does not denounce. It looks for the stakes and constraints that organize decisions.
5 – The rule of symmetry. We work as much on the decisions that failed as on those that succeeded, in order to avoid what may be called the “retrospective illusion”. This consists of thinking that somebody had foreseen an event, somebody knew about it and yet nobody did anything.
In the case of hydroxychloroquine, rule one of “generalization of significance” is of the greatest strategic importance in terms of decision-making. In qualitative studies, it is not possible to generalize a frequency with percentages. The only generalization possible when working on a few dozen people is that of the diversity of situations and practices that have been observed. This diversity is most often reduced to four or five types of case. The simple observation from a statistical point of view – probability is not my strong point – is that out of 20 people one single case is as valuable as the other 19.
In the case of the hydroxychloroquine experiment, it is possible that 20, 22 or 23 people may get better after treatment and two or three may not get better. As other doctors have pointed out, some patients heal spontaneously, so in terms of probabilities it is not possible to draw any conclusions with 24 cases. In other surveys of medical practice, I have found that it is possible in the case of rare diseases to experiment on about 20 cases and have statistically significant results. This is not a rare disease. So from a qualitative point of view it does not seem that we can draw positive or negative conclusions on the effectiveness of the treatment. The role played by chance is too high, unless a special statistical method has been used.
Should we therefore launch this treatment? Given the uncertainties in the research protocol, the policy response cannot be based on science. When the government says, “We will follow science,” this is a figure of speech, because in reality the science does not tell us what to do. It helps to inform the choices that are embedded in a concrete system of action and under constraints:
- of time – i.e. to stop the disease urgently
- logistics – this is the biggest problem in relation to masks, tests, quarantine sites and availability of beds for respiratory assistance
- of budget – for the moment the taps are fully open, following the principle that when there is a fire we don’t save on water. It may also be worth questioning this approach again by asking firemen about their experience of the fire at Notre Dame cathedral, which shows that the choices were very strategic and reasoned…
- of group norms – i.e. what is acceptable or unacceptable to say in relation to solutions of pressure groups, companies, associations, various networks, Internet, communication companies
- of culture – i.e. the usual patterns of problem-solving which are seen to be different in Asia and Europe
- of political model, which is more or less democratic or more or less authoritarian, and therefore allows more or less free circulation of information and discussions, without forgetting the “red tape” of administration which is organized for routine activities. This is the role of the red tape, and it normally works if we do not just concentrate on its failures. It is not designed for periods of crisis, which require another type of organization.
It is therefore possible to decide to invest in the “chloroquine” for human reasons, not so much health reasons, but in terms of managing anxiety and taking everyone’s concerns into account. The question of hydroxychloroquine is a matter of imagination and belief, which is quite respectable, but it is not a matter of science, since science cannot say anything about it today, at least in view of what we have been shown. If the experiment succeeds, which is humanly desirable, the Minister of Health will have been right to launch the full-scale experiment. If the experiment fails it will be unfair to blame him, but he will probably be used as a scapegoat.
Science can help to reason better, in a less emotional, but still sensitive way, and with room for doubt. Paradoxically, in some ways it is as much about scientific behavior, with a small s, as it is about scientific results that can help us today to make better political or personal decisions. It is this behavior that can help us not to go back to the way things were before, by understanding the constraints of action in order to better remove them and change them.
PS I have been reading Jancovici for a very long time on energy problems, since it is one of my anthropological specialties. I am paying particular attention now, as he is putting on Linkedin a whole series of reflections which show how what we are experiencing today is a laboratory experiment to reveal the dependent and independent variables, on a macrosocial scale, which we must regulate in order to fight against global warming. The exceptional recession we are experiencing today, which some call degrowth, highlights the difficulty of combating greenhouse gases in terms of production, distribution, consumption and recycling outside a period of crisis. It is worth the cost of reading it.
Paris on March 24, 2020