Re: Unicef data

3 views
Skip to first unread message

MG Subramanian

unread,
Mar 6, 2012, 4:02:54 AM3/6/12
to bangalore-mo...@googlegroups.com
How does one take a complex behaviour, "the tendency of teachers to award corporal punishment?" into a model One can possibly lump it along with some quantification of lumped with social inhibitors. But the moment you "lump" something you lose sight of its detailed characteristics" its specific causal mechanisms including any relation it might have to other social beliefs. For example, the Unicef article, interestingly, mentions the "social distance" between teachers and students as one enabling factor( term mine,not the article's.) for corporal punishment. How does one study such things in quantitative "model thinking" terms?

Let me deflect a bit and add cheer to this post which has been without it so far!

I was at an education meet yesterday organised by Sikshana.Mr. ES Ramamurthy, the founder of Sikshana, a very insightful gentlemen and leader who has been working to improve the Govt Schools was present. Read his article here.One of principles on which  ESR Sikshana  has been working on is that " each school"  is unique in terms of what is needed to improve the learning outcomes. That is to say a million schools can be improved by one school improving at a time! Sikshana has been able to improve performance in schools by doing such simple things as holding school days at a handsome expense of Rs.2000.

This may appear to be not such good news for a group that is anxious to model (and find silver bullets?) for something as complex as the retention patterns in schools.The "one school at a time" attack strategy does argue for communities taking greater ownership of the schools that serve them. How the attitude of communities taking ownership for community problems can be nucleated and spread itself is a worthwhile model to study.You can read an interesting example here that, in my mind can be modelled quantitatatively for a better understanding of the kind of qualitative and imagninative thinking needed. In many similar models there are tipping points beyond which a positive reinforcement cycle takes off. Sucessful reforms depend on the skill with which a reformer identifies the true leverage that can trigger off positive behaviours.

I hope that the next section "Modelling people" can will give us some fresh tools...if it only presents thorugh another lens light what the leader responsible for this knew!

-ganu


.




On Tue, Mar 6, 2012 at 12:08 PM, Anoop <anoopka...@gmail.com> wrote:
We should also consider the effect of Corporal punishment on the
number of school drop-outs. According to a September 2011 UNICEF
report, 74% of all dropouts do to the fear of physical, mental or
verbal abuse in a classroom. I could not find the report itself but
this was mentioned in an Open page article in 'The Hindu' --
http://www.thehindu.com/opinion/open-page/article2932688.ece

Reply all
Reply to author
Forward
0 new messages