In these difficult times of infections and deaths due to the Covid-19 crisis, I have been hard put to gain accurate information on numbers for places such as Dharavi in Mumbai. Quite often, there are conflicting figures emanating from different sources for the same day’s data.
This is symptomatic of a larger chronic issue that is affecting India for quite some time. This data-ignorance is getting exposed during the coronavirus pandemic more than ever.
Indians have become alarmingly indifferent — and often inefficient — when it comes to dealing with data and its use. It is ironic that this is happening in the land where Chanakya — much before the time of Christ — pioneered the use of statistical analysis for better governance. One would thus have expected that India would be better able to model the global data on Covid-19.
Dharavi and BCG vaccination
Take this interesting data-based insight that has emerged from the Texas Medical Centre in Houston on the correlation between low death rates and nations that have had a vigorous Bacille Calmette-Guérin — BCG — vaccination programme.
Given that India has had mandatory BCG vaccinations since 1978, it seems a given that this study should have been undertaken further to check the situation in a closed environment — such as Dharavi — to see if the correlation gets any stronger.
Unfortunately, I have not seen any such study. In fact, I have not seen any data-based analysis done in a meaningful manner for any aspect of the pandemic in India. A great deal can be achieved by examining the data via sources that are freely available on the web. As per my understanding, Dharavi has had declining numbers of infections and declining rates of infections for the past few days. According to my calculations, the average rate of increase of infections for the past week has been 5.7 per cent and it has been hovering around 3 per cent for the last three days. This data is in stark contrast with what was happening a few days ago when Dharavi clocked a weekly average infection rate of 9.6 per cent.
In fact, the projections for Maharashtra are also definitely getting better. The hype is all about the spike in the absolute numbers of daily infections in Maharashtra. What we are missing is that the rate of infections has been steadily declining.
The average rate of increase in infections in the state in the past week, as per my calculation, is 7 per cent. On the other hand the average for a week, in the middle of April, stood at nearly 21 per cent.
Keeping in mind the model proposed by the eminent Israeli scientist Isaac Ben Israel, it seems safe to predict that Dharavi shall be out of the woods by the end of May and Maharashtra by 10 June. Ben Israel suggests — through very rigorous data analysis — that the coronavirus will lose steam in about 70 days. One must remember that these inferences are not mathematical laws like the law of gravitation but are reasonably accurate approximations.
Along these lines, it would have been very productive to examine if Dharavi has a vigorous BCG vaccination programme. Given its very low death rates and its low numbers of infections, this aspect needs close statistical examination.
Absence of accurate data
There is one other statistical opportunity that Dharavi offers but once again I see no effort to undertake this exercise. Given that the coronavirus infects a significant number of individuals who remain asymptomatic, a simple and effective random experiment can be undertaken to determine the true numbers of those infected in Dharavi. One could then extrapolate this data for the entire country. The other advantage that accrues in the context of the crisis is that one can use data-based inferences for devising strategies and also help alleviate public anxiety.
The disappointing part is that these matters would have been very interesting projects for school and college students who are worried about their studies and are spending time at home. But the lack of connection between what students are taught and what is going on in the real world is worrisome.
Nothing bears this out with greater clarity than my recent interaction with a bunch of college teachers — drawn from the discipline of mathematics — for an online seminar on some facets of the Covid-19 crisis. This worthy group turned out to be clueless on devising an experiment to estimate their chances of contracting the coronavirus. However, the problem was in the realm of elementary data analysis and within the grasp of a grade nine student.
I have also been searching for data related to the Covid-19 cases for different regions of India. It has become an exercise in confusion, since very often, different sources offer different figures for the same situation. It is hard to find a one point authentic source for aspects of data.
It seems that the value of data and data-related information is lost on us as a nation. Such a situation is rooted in the chronic inability of India’s educational institutions to offer any worthwhile training in basic data analysis. This has resulted in a culture that abhors the value of data and shuns any engagement with it. What sort of harm is this causing the nation? All our institutions that deal with disciplines such as medicine, the life sciences, the social sciences and the humanities have very little to do with any training in data or in quantitative thinking. This begins at the level of schools and runs right through almost all educational institutions, barring a very small number of exceptions.
Data training in universities
Take a venerable institution like the All India Institute of Medical Sciences — AIIMS. As far as I can tell, it does not have any worthwhile statistics or probability theory experts. The same is true of the much-vaunted National Institute of Immunology. Contrast this with the highly regarded MD Anderson Cancer Centre at Houston, Texas, which has as many statisticians, probability and mathematics experts, as the number of medical scientists.
How else can one understand genome data and genome sequencing? And without these, there is not much one can do when it comes to understanding the genetic basis of disease and therapy. For instance, genome sequencing is proving to be advantageous in treating newborns and infants for various diseases. It is just as useful in drug trials and pharmacogenomics. The list goes on.
What many of us in India do not realise is that the same data analysis that works for genome research, works just as well for financial data or for myriad of other needs. I am ready to forego such ambitious goals for most of our young students. Instead, I just ask for very basic training in data such as what Delhi University had mandated in 2013 for undergraduates under the Four Year Undergraduate Programme, regardless of their disciplines of study. The course was very practical in nature and quite useful for the students in all kinds of situations and for their careers. It also proved to be very popular as well as effective in creating a new culture on campus. For instance, students of literature were also gaining new insights into language and literature through it.
Unfortunately, it was taken off the menu when the entire package of reforms was rolled back without any reason by the University Grants Commission (UGC), under the influence of the HRD ministry in June 2014. This thoughtless action caused incalculable harm to the nation and set us back by many, many, years.
A few days ago, Amitabh Bachchan tweeted with delight what he thought was an amazing fact, peculiar only to the year 2020. He pointed out that if one were to add her/his birth year to one’s age, she/he would end up with 2020. It is painfully obvious that Bachchan has overlooked a simple arithmetical truth that renders his observation valid for each calendar year and not just for 2020. What is even more painful is that within a short span of his tweet, there were more than fifty thousand likes and several thousands of re-tweets. Thousands of Indians had missed a simple arithmetical flaw in his assertion. I rest my case.
The author is the former vice-chancellor of the University of Delhi, a distinguished mathematician and an educationist. Views are personal.