NUMBERS MATTER: How statistical misinterpretation can lead to unecessary Alarm

Posted by Ishmeet Nagpal
June 21, 2017

Self-Published

 

‘3 Mn Girls Lost To Infanticide In 2011: Exposing India’s Horrid Gender Bias’ is the title of an

article published by Mr.Abhishek Jha in Youth ki Awaaz. And this article was being used for

crowd sourcing for a social media campaign on protecting the Girl Child. Being active in this

field for more than a decade this data flustered me…I knew the data was wrong but I wondered

how did the author write so confidently about it. This made Dr. Ishmeet Nagpal and me go searching for the

source of the data. We had our Eureca moment when we found a paragraph in ‘CHILDREN IN

INDIA 2012 – A Statistical Appraisal, Social Statistics Division, Central Statistics Office,

Ministry of statistics and Programme Implementation Government of India. Obviously a very

reliable data source. The paragraph said:

“While an absolute increase of 181 million in the country’s population has been recorded

during the decade 2001-2011, there is a reduction of 5.05 millions in the population of

children aged 0-6 years during this period. The decline in male children is 2.06 million and

in female children is 2.99 millions”.

The above data refers to the decline in the population of children mainly due to decline in

fertility rate and does not refer to infanticide or pre-birth sex selection in particular. While

NCRB puts the total cases (male and female) of Infanticide at 1758, the estimates of missing girls due to pre-

birth sex selection is put around 6-7 million by various sources.

There was yet another campaign designed on IMR which had communication proclaiming

that 60% of children die before the age of 1 year. Luckily it was shared with me before

going public and I had to point out to them that it is 60 per 1000 children born alive and not

a percentage.

In both the instances the mis-interpreted data would have made a huge impact on the

readers causing alarm. As responsible communicators is it not the responsibility of those

designing major campaigns to cross check data? How easy it is to call someone from the

field and take a feed back and how seldom it is done. Data cannot be understood without

understanding the logic behind it.

I am often asked to comment on improvement in sex ratios at ward level and newspapers at

time carry stories saying that a particular ward has shown great improvement in sex ratios

or has gone down drastically. I often ask them to check the number of births in the ward. If

the births are low say below 3000, even a small change in the number of girls born will

show up as a major change in sex ratio as sex ratios are calculated as number of girls per

1000 boys!!

Similarly increase in number of abortions does not necessarily reflect increase in pre-birth sex

selection nor does increase in teen-abortions indicate promiscuousness. Data on increase in the

number of atrocities on women does not necessarily reflect a worsening safety situation. It could

be due to increased reporting or changes in the definition of the indicator across yeas and

sources.

Data is important to support our writing and to convince our readers about the importance of a

particular issue. However, it is absolutely important that the data is interpreted, understood and

presented in the right way. When in doubt, call an expert from the field to interpret the data for

you. Please don’t mis-inform and mislead the readers.

 

Authored by :

Dr. A L Sharada

Director

Population First

Youth Ki Awaaz is an open platform where anybody can publish. This post does not necessarily represent the platform's views and opinions.