Research Ethics, Whistle Blower & Tuskegee Experiment

What was done cannot be undone. But we can end the silence. We can stop turning our heads away. We can look at you in the eye and finally say on behalf of the American people, what the United States government did was shameful, and I am sorry.

–          Bill Clinton, May 16, 1997

In 1932 U.S Public Health Service at Tuskegee Institute in Tuskegee, Alabama, conducted a study to determine effect of untreated syphilis in black men. It was supposed to be a six months project. But what happened in reality was not just unethical but inhuman.

Instead of six months the study continued for 40 years—from 1932 to 1972. The study included 600 black men, 399 with syphilis and a control group of 201 who did not have the disease. The men in the study were the sons and grandsons of slaves. Most had never been seen by a doctor. The researchers made announcements in churches and cotton fields for participation, many volunteered thinking that they would free medical treatment.

tusk 2

The men in the study were never told that they had syphilis, a sexually transmitted disease. Instead, government doctors told the men they had “bad blood,” a term that was commonly used to describe a wide range of unspecified maladies.

In the mid-1940s, when penicillin became the standard cure for syphilis, the Tuskegee subjects were not given the drug. Even as some men went blind and insane from advanced (tertiary) syphilis, the government doctors withheld treatment and prevented subjects from getting treatment from other doctors in the region as their primary interest was in observing advancement of disease and then doing autopsy. To ensure that the families would agree to this final procedure, the government offered them burial insurance—of $50—to cover the cost of a casket and grave.

tuske 1

One interesting fact about this experiment was role played by a black nurse called Eunice Rivers who was part of these experiments from beginning to end i.e. for 40 years and was actively involved in getting volunteers for experiment and preventing them from getting treated by other doctors in the region. Later a film was made on these experiments called – Miss Ever’s Boys.

evers boys

The research project was finally stopped after Peter Buxtun, a former employee of PHS, blew whistle and shared the truth about the study’s unethical methods with a journalist Jean Heller. On July 25, 1972, the truth was published in newspapers; it resulted in public outcry that ultimately brought the notorious experiment to an end.

tusk 3

This episode resulted in strengthening guidelines for protection of human subjects in research. Fred Gary, a civil rights attorney, filed a class-action lawsuit on behalf of the men that resulted in a $10 million out-of-court settlement for the victims, their families, and their heirs.

“I want to live in peace and harmony. How can we love the Lord, whom we’ve never seen, and hate our fellow men, whom we see every day? I want to get along.”

–          Herman Shaw, survivor of Tuskegee Experiment.

Finally in 1997, Bill Clinton apologised to survivors of Tuskegee experiments.

herman shaw

Fable, Statistics, Type I & Type II errors

In statistics while testing a hypothesis, you are bound to make type I error (denoted by alpha) or type II error (denoted by beta). Both are inversely proportional to each other i.e. if you try to reduce one type of error, you increase the other type of error.

Type I occurs when you reject null hypothesis when it is true i.e. a blood test will show that you have disease when in fact you don’t. Type II is opposite, here blood test will not show disease, when in fact you have one.

images

Many found this confusing . I found an interesting fable which helps you to remember the difference between two types of errors.

A beautiful maiden married a handsome prince called Alpha. But after marriage she found that he was impotent. So she got into relationship with Alpha’s ugly looking brother Beta. So for subjects of kingdom there was relationship between Alpha and maiden, when in fact there was none, while according to subjects there was no relationship between Beta and maiden, when in fact there was one.

The King was also a statistician forgave maiden for Type I & Type II error.

 

Game Theory, Colonel Blotto and Gerrymandering

All men can see these tactics whereby I conquer, but what none can see is the strategy out of which victory is evolved.

-Sun Tzu

In game theory there is a classical war game called Colonel Blotto. In this game there are two players and the player who devotes most resources to a battlefield wins that battle, and the gain (or payoff) is then equal to the total number of battlefields won.

There are certain rules of the game

  1. In battlefield the player that has allocated the most soldiers will win.
  2. Both players do not know how many soldiers the opponent will allocate to each battlefield
  3. Both players seek to maximize the number of battlefields they expect to win.

Let us take an example. There are two players A & B. Each has 100 soldiers and they have to capture 10 forts (or 10 battle fields), now each player decides how many soldiers to send for each fort ex. suppose A send 10 soldiers to capture fort # 1 while B sends 9 soldiers, then A wins the fort. If both send 10 each then there is a tie. Idea is to win any many forts as possible and one who wins maximum number of forts (or battles) wins the war.

war strategy

Since you don’t know what opposing player will do, there can be lot of combinations. I am not going statistics part of it, there are lots of books on game theory/operations research which will tell you how to come up with optimal solution/Nash equilibrium.

One strategy could be, since each player has 100 soldiers and 10 battles to be won, so each send 10 soldiers per battlefield, but none wins, since it will end in tie.

Other strategy could be A continues with earlier strategy of – “10 10 10 10 10 10 10 10 10 10”, while B comes up with another strategy of “1 11 11 11 11 11 11 11 11 11”. B will lose one battle (1:10) but win other 9 battles (11:10), therefore win the war.

war

This theory has lot of applications esp. in politics. Each party has limited resources, so sends most resources in areas where chances of winning are high. But sometimes politicians alter the boundaries of district/state in such a manner that altered district has maximum number of supporters and victory is assured. It is called Gerrymandering.

640px-The_Gerry-Mander_Edit[1]

There is interesting story on origin of this word. The word was created in reaction to a redrawing of Massachusetts Congressional election districts under the then-governor Elbridge Gerry. In 1812, Governor Gerry signed a bill that redistricted Massachusetts to benefit his Democratic-Republican Party. When mapped, one of the contorted districts in the Boston area was said to resemble the shape of a salamander. Hence the word is combination of Gerry+ Salamander= Gerrymander.

Problem Solving, Working Backwards and Pirate Game

“If I had an hour to solve a problem I’d spend 55 minutes thinking about the problem and 5 minutes thinking about solutions.”

― Albert Einstein

Psychologists have done lot of research in area of problem solving. They have come up with number of methods like hill climbing (reach one goal at a time, then go for next one till you reach the final goal), mean-end analysis (identify current state and end state, now work out what needs to be done to reach final state), forward chaining (identify the givens and work directly towards goal), considering analogous problem (used in Tata Innovista, where idea from one industry is used to solve similar problem in another industry ex. technique of drying of plaster used by Tata Projects was used by Titan to reduce cycle time for drying objects post gold plating) etc.

One of the techniques is to solve problem working backwards i.e. working backwards from final goal through all steps necessary to reach that goal.

Game theory has one such interesting game called as Pirate Game, which when solved backwards can give an interesting and unexpected result.

Game is like this. Suppose there are 5 pirates (it can be any number) and they find chest with 100 gold coins, now pirates have to distribute it among themselves. Solution is not as simple as each taking 20 coins each, because pirates have their own laws and traditions and distribution has to be done as per traditions.

pirates[1]

Let us give number to pirates starting from 1 to 5, with 1 being senior most and 5 being junior most.

As per tradition, senior most pirate will propose distribution. The pirates, including the proposer, then vote on whether to accept this distribution. In case of a tie vote the proposer has the casting vote (i.e. at least half of them should support proposer). If the distribution is accepted, the coins are disbursed and the game ends. If not, the proposer is thrown overboard from the pirate ship and dies, and the next most senior pirate makes a new proposal and game starts again.

Now as per game theory, we have to assume that all pirates are rational; hence will base their decisions on three factors.

  1. Each pirate wants to survive.
  2. Each pirate wants to maximize the number of gold coins he receives.
  3. Each pirate would prefer to throw another overboard.

coins

Game will start with pirate #1 as proposer (being senior most), we may conclude that pirate #1 will allocate little for himself for fear of being voted out and thrown in sea. But actual solution is quite different from what we are assuming.

Actual solution( distribution) is shown below.

#1: 98 coins

#2: 0 coins

#3: 1 coin

#4: 0 coins

#5: 1 coin

So #1 gives away only 2 coins, keeps 98 coins and manages to survive.

For solution we need to work backward, suppose #1,#2 & #3 are thrown overboard and only #4 and #5 remain, here# 4 will take all 100 coins ( he has casting vote in case of tie) and given #5 nothing.

Now if #3,#4 & #5 are left, #3 knows very well that if he is thrown overboard, #5 gets nothing, so he offers one coin to #5 and gets his vote and gives #4 nothing, so #3 gets 99 coins , #4 nothing and #5 one coin.

If #2, #3, #4 & #5 remain, #2 being a rational player knows what will happen if he is thrown overboard. To avoid being thrown overboard, he can simply offer one coin to #4 and gets his vote. Because he has the casting vote, the support from #4 is sufficient. Thus he proposes himself 99 coins, #3 nothing, # 4 one coin and #5 nothing.

Since #1 is a rational player, hence knows all these things, so he gets votes from #3 and #5 by giving them one coin each and gives #2 and #4 nothing and keeps remaining 98 coins with himself.

The game can be played for n number of players; the pattern of distribution is as follows.

No of Pirates (n)              No of coins got by each (nth, n-1 th …)

1                                                      100

2                                                            100, 0

3                                                           99, 0, 1

4                                                            99, 0, 1, 0

5                                                            98, 0, 1, 0, 1

6                                                           98, 0, 1, 0, 1, 0

7                                                 97, 0, 1, 0, 1, 0, 1

 

Game Theory, Diner’s Dilemma and Welfare Schemes of Government

Treat each federal dollar as if it was hard earned; it was – by a taxpayer.

–          Donald Rumsfeld

In Game theory there is interesting game called as Diner’s Dilemma.Diner’s Dilemma is a game-theory situation with several players. A diner’s dilemma occurs when several participants attempt to obtain the highest possible personal reward, but instead find themselves in an unfavourable situation.

imagesCAB18N3P

If five friends were asked to go to a restaurant individually and pay their own bill, they will eat dishes that are cheaper. Now if all of them decide to go to restaurant together and agree to split the bill before ordering dishes, assumption here is per head cost will be lower than what they would have paid had they gone alone. This is where Diner’s Dilemma sets in. Each one of them thinks that since bill will be split i.e. cost will get distributed between other members, he/she will order an expensive dish, so everyone ends up ordering dishes more expensive than what he/she would normally buy,  and end up spending more money i.e. they all end up facing the outcome they tried to avoid- a more expensive meal.

cartoon

There is one more variation of this game; if the same group was told that bill will be paid by someone else, the group will eat more than what they would eat normally and order most expensive dishes.

Our government, which believes in social welfare, can learn lot from this esp. the way it wastes tax payer’s money on several social welfare schemes. As long as people pay for services they use, they will use it reasonably, but once politicians decide to subsidise or distribute free of cost services like water, electricity, fuel, fertilizers etc., consumption will go up, wastage will increase and tax payer will end up paying even more.

waste money

Anthropology, Regression and Dunbar’s number

“The number 150 really refers to those people with whom you have a personalised relationship, one that is reciprocal …if you asked them to do a favour, they would be more likely to say yes than those outside the 150.”

–          Robin Dunbar in his book  “How Many Friends Does One Person Need?”

Robin Dunbar, a British Anthropologist, did interesting research in 1992. He studied 38 species of primates (monkeys, apes etc.); he calculated size of neocortex to rest of brain and size of social group for each of them. He found positive correlation between two i.e. higher the neocortex ratio, larger the size of social group.

robin

Based on values of these two variables i.e. ratio of neocortex to rest of brain and size of social group, he developed regression equation, he then calculated ratio of neocortex to rest of brain for humans and used regression equation to find corresponding value for size of social group. The value obtained was 148, which was rounded off to 150. In other words, extrapolating from the results of primates, humans can only comfortably maintain 150 stable relationships. This number (i.e.150) was called Dunbar’s number.

dunbar graph

Next Dunbar and his team decided to gather empirical evidence to support their statistical analysis. They estimated sizes of a Neolithic farming village, Hutterite settlements, basic unit of professional armies and organisational unit. The number was around 150, which supported statistical findings.

hut

Research study has shown that Dunbar’s number is also applicable to online social networks. So ideally you should have around 150 connections on Facebook!

What if you want to maintain stable relationship with more than 150 people? You will have to increase size of your brain.

Occum’s razor, Pakistan and Jogendra Nath Mandal

“The principal objectives that prompted me to work in co-operation with the Muslim League was, first that the economic interests of the Muslims in Bengal were generally identical with those of the Scheduled Castes…and secondly that the Scheduled Castes and the Muslims were both educationally backward.”

–          Jogendra Nath Mandal in his resignation letter to Liaquat Ali Khan

William of Occum, a British philosopher came up with a problem solving principle, which was named after him- Occum’s razor. What it states is if there are multiple explanations for a phenomenon, explanation with least assumptions should be selected. In other words go for simplest explanation.

2088_c953_512

There are many books which explain what lead to partition on India. They analyse a lot on what happened between 1940 to 1947, and end up blaming Mahatma Gandhi, Nehru or Jinnah. Many leftist historians use Occum’s razor to explain partition. India is depicted as upper caste Hindu dominated state, indirectly ruled by zamindars and industrialists, which endangered economic, political and religious freedom of minorities, lower castes and tribals, and Pakistan was outcome of this.

partition

But Occum’s razor can also lead to wrong conclusions, because at times complexity cannot be ignored. Jogendra Nath Mandal was leader of scheduled castes in Bengal, he was quite influential in politics of Bengal, infact he managed to get Dr. Babasaheb Ambedkar elected to constituent assembly of India from Bengal.

mandal

He was also one of the founding fathers of Pakistan, and member of their constituent assembly. He became first law and labour minister of Pakistan. In Bengal, most of the landowners were upper caste Hindus while Muslims and Scheduled Castes were cultivators. He believed that, Scheduled Castes will have equality (as caste Hindus will cease to dominate) and freedom from oppression of Hindu landlords and money lenders (who also happened to be caste Hindus); besides M.A. Jinnah had assured them freedom to practice religion in his speech of 11th August 1947.

jinnah

“You are free; you are free to go to your temples. You are free to go to your mosques or to any other places of worship in this State of Pakistan. You may belong to any religion, caste or creed—that has nothing to do with the business of the State.”

― Muhammad Ali Jinnah

But Jogendra Nath Mandal soon realised reality was complex. First, his demand to have two more Scheduled Caste members as ministers was ignored by Liaquat Ali Khan, then Prime minister of Pakistan. Secondly, Hindu- Muslim riots made him realise that, post Jinnah, Pakistan was very different from what he had thought. In 1950 he resigned from cabinet and returned to India.

“It is with a heavy heart and a sense of utter frustration at the failure of my lifelong mission to uplift the backward Hindu masses of East Bengal that I feel compelled to tender resignation of my membership of your cabinet.”

–          Jogendra Nath Mandal in his resignation letter to Liaquat Ali Khan