Statistics, Survey methodology, Computing in R, General management practices, Sanskrit, ...
Ann Arbor, MI
Joined April 2009
- Tweets187
- Following45
- Followers79
- Likes60
Sharan Sharma retweeted
How is ageing reshaping systems of support in India?
Drawing on NCAER IHDS data, new research explores how family structure, functional health, labour force participation, and state support interact to shape the wellbeing of Indians aged 60 and above. The findings point to declining intergenerational co-residence and an expanding role for pensions and health insurance, highlighting an evolving balance between family, individual resources, and state support in ensuring wellbeing and security in later life.
@DebasisBarik25 | @Neerad_D | @snsharma | @SonaldeDesai | @ihdscorner
Read more: bit.ly/4fEdmzr
Sharan Sharma retweeted
📜Paper alert: bit.ly/3UPAEqc
@snsharma @SonaldeDesai @DebasisBarik25 & Om Prakash Sharma use @ncaer @ihdscorner data to understand omitted births in fertility histories—conclude they are 'non-ignorable' & non-random; estimate prevalence at 4%; & discuss its predictors.
Sharan Sharma retweeted
@snsharma & Michael Elliott develop a method of detecting survey falsification. The method uses multilevel models & they illustrate its use in a panel survey in India measuring TV ratings bit.ly/33VPz5Q
Sharan Sharma retweeted
Some of the TV panels discussing exit polls are larger than the sample size :D
If you wanted some real case-studies on monitoring interviewers in challenging situations. Also contains a fair bit on detecting falsification.
Zeina Mneimneh, @snsharma et al. contributed Case Studies on Monitoring Interviewer Behavior in International and Multinational Surveys bit.ly/2ycBnqT 15/16
Sharan Sharma retweeted
Just loving this massive portrait at Bengaluru Airport made entirely from dice to commemorate the life of the famous Indian Mathematician Srinivasa Ramanujan. – at Bengaluru, India
Super story to illustrate very clever use of technology and modeling - sadly, to defraud.
wired.com/2017/02/russians-e…
Apparently, a few decades ago some grant reviewers didn't like the concept of using behavior coding for question evaluation. Slide 7 in this ppt: bit.ly/FowlerCoding #SurveyMethodology
1
Love it when technical books are interspersed with humor. This is from: "Measurement Error in Nonlinear Models", Carroll, Rupert, and Stefanski (2006). #Statistics
Q: Why should I not trust quarterly numbers from the Central Statistical Organization (CSO)?
A: Because my dad founded CSO. #StrongReasoning
indianexpress.com/article/in…
Nice preface in #Statistics (Freedman, Pisani, and Purves, 1998). But a bit surprised at the highlighted phrase.
(2/2) i.e. despite energy drain, may be publicly valuable to call out BS emanating from a celebrity vs. arguing online with 'agentsky9991'.
Wish to propose a weighted version of the "Bullshit Asymmetry Principle", where weights = "importance of perpetrator" (1/2)
Heartening to see #Sanskrit like this. Lovely way to keep ancient literature "by heart":
timesofindia.indiatimes.com/…
Careful while reading the output from 'summary' in #rstats.
?summary:
summary(object, ..., digits = max(3, getOption("digits")-3))
