Monday, November 23, 2015

Search vs Hunt in Websites

I wasted forty minutes at a bank’s website last week. I needed to recharge a prepaid card. At login, there is a choice between the savings account login and prepaid card login. Which way do I go? The bank’s approach is to let the customer find out by trial and error. I tried the prepaid card login and discovered I could not do a recharge there. I found out the customer care number and called them. The support “executive” had no clue either. He told me that at login I should choose the prepaid card option. I told him that is exactly what I had tried, without any use. He told me to hold on and went off to get an answer. This business of consulting someone else took place three times and I had to hold on each time.
The first time he came back with borrowed wisdom, he asked me to log into my savings bank account. I did, and then had a question. Which tab do I select? He suggested that I go to “funds transfer” – a good guess, as I could perhaps transfer some money to my prepaid card as I transfer money to any other card. However, that did not work, so he went off to consult his colleague again. “There is a tab named Cards”, he said, “Click on that!” After bumbling around like this for quite some time, we finally discovered how to recharge the stupid card. He had been kind enough to stay on the line, suffering through all this with me. For the bank, this was not wasted employee time – it was probably free training provided by the customer!

This raises the question of the point and click model to tell a system what you want. It is very convenient, like a menu in some countries with photographs to guide the tourist who does not share a language with the staff. You find an item you recognize and point it out to the waiter; you get survive another day! This point and click model does not work when you have go through a hierarchical maze of tabs, making selection after selection without any clue till you get to see what you are looking for. In that situation, the “selection by search string” model works far better. For instance, Windows 10 has a slot at the bottom of the desktop labelled “Search the web and Windows”. If you type in “Hide rows in Excel sheet” and it takes you to your search engine, which gets you the answer – it may come from Microsoft websites or from someone else’s. You get a set of possible answers and can examine and select one. A great feature of this is that you are not restricted to a single word to say what you are looking for; many “help systems” force you to work under this restriction. It is like working with one hand strapped to the chair’s armrest! With a multi-word search string, you usually get the answer you are looking for in one shot.

How can a website, say a bank’s website, provide for access through a search string? Can the website let me click “Help” and type any of the search strings I list below as examples?
         Recharge prepaid card
         Register a new beneficiary for interbank funds transfer
         Form for requesting closure of account
         List of bank holidays in Delhi

I do not deny that the graphical user interface (GUI) involving icons, and point and click was a great invention; so is access to a webpage through the use of a multi-word search string. Like a screwdriver and a hammer, they are complementary. One does not replace the other.

HCI: Comment No. 4
= = = = =

Friday, November 6, 2015

Railways and the Swachh Bharat Abhiyan


There has been considerable concern in India about the fact that train toilets spread solid waste and urine on railway lines all over India, creating a health hazard. News reports indicate that in 2015, we are just beginning the effort to reduce this menace, installing new types of toilets on selected trains. At this rate, it would take many years to reduce public health risks caused by rainwater carrying contaminating material to village wells and ponds, spreading a variety of diseases. No estimates seem to have been made of the number of deaths caused per year due to this. Medical colleges should carry out fieldwork and research about water-borne diseases spread by this contamination of drinking water.

Sushmi Dey writing in the Times of India, Nov 13, 2014
( India loses most kids to diarrhoea, pneumonia: Study )
reports that “India tops the global list with 318 deaths per 1,000 children under five years of age due to pneumonia and diarrhoea“.

Diarrhoea is the third most common cause of death in under-five children, responsible for 13% deaths in this age group, killing an estimated 300,000 children in India each year, according to
Causes of neonatal and child mortality in India: nationally representative mortality survey

Undoubtedly, there has been a reduction in under-five deaths during the last decade, but reports indicate that India will miss the millennial goal in this regard. Visit Business Standard, March 7, 2015:
Millenium Development Goals: India's progress so far

It is surely worth examining the causes of diarrhea deaths among Indian children. It would be very surprising if poor quality of drinking water is not a significant factor involved.
Science colleges, particularly those in rural areas, should acquire capabilities to test the quality of drinking water used in their neighbourhoods. It would be tragic if a degree in chemistry and biology does not equip a student to test if water is fit enough to drink! We must train students to carry out these tests, and then encourage and support them in doing relevant fieldwork to utilize their training for the benefit of people in the area. An interesting paper shows the results of testing drinking water in colleges in Kolhapur:
We should compliment the researchers at an engineering college who carried out this research. 

The big risks are in bacterial and viral contamination of water sources. Chemical testing alone will not be sufficient. The following website provides some relevant information for those interested in testing water from this point of view. 
http://www.water-research.net/index.php/water-testing/bacteria-testing/coliform-bacteriaA web-search shows a number of vendors who sell water-quality-testing kits.
Railways have their own responsibility in this regard. They cannot magically transform all train toilets to safe toilets. However, they can consider interim measures to ameliorate the problem. They could use small bogies attached to trains to carry a few hundred litres of a solution to spray the space between rails whenever the train is traveling above a set speed. This will avoid spraying too close to railway stations. Experts would have to select cost-effective solutions for the purpose. There are many options to select from, to kill germs and parasites in solid waste before rainwater carries them into village water sources.


A significant initiative by railways in this direction would be a major contribution to the Swachh Bharat Abhiyan.