Saturday, August 28, 2010

Questioning the infallibility of solutions provided by GMAC for GMAT practice problems

I've largely liked The Official Guide for GMAT® Review, 12th Edition. It's a high-quality, well-built book. A trustworthy book. Makes you feel that when GMAC says that 1+1=10, it's gotta be true. So is the case with - a well-built and trustworthy website.

Except that I've come across a very small number of questions, the GMAC-provided solutions to which have shook my faith in the infallibility of GMAC's solutions.

Example 1: I was horrified when I saw that GMAC has provided the following essay as a model essay - receiving the highest rating - for an Analysis of an Argument question, considering the following three statements about the Argument essay:
  • "In this task, you are not asked to state your opinion but rather to analyze the one given." - Page 759, The Official Guide for GMAT® Review, 12th Edition
  • "Perhaps the most serious error one can make on the Argument essay is to disagree with the author's conclusion." - Page 534, Kaplan GMAT 2010-2011 Premier
  • "Violate the essay directions, either by not taking a position on an Issue or by taking one in an Argument, and you risk scoring a 2." - Page 539, Kaplan GMAT 2010-2011 Premier

Example 2: Among the GMAT Teaser questions on, QuestionID=25 succeeds in teasing me. According to GMAC, "The fourth choice is correct. It can be inferred from information given in the third and fourth sentences of the second paragraph: In addition, any institution that holds 20 percent or more . . . the value of the stock would plummet.". I say BS. The passage never says that the stock must plummet. It only indicates that the stock would plummet unless there's an explanation that can be provided.

Friday, August 27, 2010

A coincidence? Perhaps. But intriguing.

Today morning I was thinking about nondestructive testing in the context of civil airliners. I thought, it's important to nondestructively test each aircraft before it's delivered to its customer. That was about it. I forgot about it for the next few hours.

A few hours back, I conducted some queries related to ETS, GRE, GMAC and GMAT, and hit upon an interesting-looking article.

To my surprise, the Google-served ad below the article read Non-Destructive Testing, and linked to a company providing nondestructive testing systems.

How's it possible? I clearly remember that at least in the last 1 year, I haven't conducted any query related to nondestructive testing.

Thursday, August 26, 2010

A now-funny "clarity" about the meaning of Internet, in a Dec'94 article in NYT

"The E.T.S. warned that anyone releasing questions without permission would face charges of copyright infringement. Ms. Cole said testing service officials did not think that people were leaking questions from the computerized tests. She said officials have been monitoring computer conversations between college students on the Internet, the huge computer connection a network allowing communication by computer, to see if college students talked about cheating on the G.R.E., or whether questions were being passed via computer conversations." - Computer Admissions Test Found to Be Ripe for Abuse, NYT, Dec'94

It feels so good when I recollect one of the faint memories from class III or IV, when there was no Internet here, when NYT had to add words clarifying the meaning of Internet.

Does GMAC reveal the identity of an examinee to the readers of the examinee's AWA responses?

"Readers are trained to be sensitive and fair in evaluating the responses of nonnative speakers of English." - Page 760, The Official Guide for GMAT® Review, 12th Edition

Does this imply that the nationality of a test-taker is revealed to the readers who evaluate a test-taker's responses in the AWA section (of the GMAT)? If yes, is this disclosure leading to the readers awarding ingenuine scores to nonnative speakers of English?

Monday, August 23, 2010

Apple's "Thoughts on Flash" are hypocritical

"3.3.1 — Applications may only use Documented APIs in the manner prescribed by Apple and must not use or call any private APIs. Applications must be originally written in Objective-C, C, C++, or JavaScript as executed by the iPhone OS WebKit engine, and only code written in C, C++, and Objective-C may compile and directly link against the Documented APIs (e.g., Applications that link to Documented APIs through an intermediary translation or compatibility layer or tool are prohibited)." - iPhone Developer Program License Agreement

"We know from painful experience that letting a third party layer of software come between the platform and the developer ultimately results in sub-standard apps and hinders the enhancement and progress of the platform." - Thoughts on Flash, Steve Jobs, Apr'10 [link 2]

My biggest, and in fact the only problem with Apple forbidding Applications that link to Documented APIs through an intermediary translation or compatibility layer or tool is that it allows - hypocritically - JavaScript and Safari on to the iPad/iPhone/iPod touch. Safari is a native application that links to Documented APIs. And it runs JavaScript applications on top of its native code.

If JavaScript can be allowed inside Safari, a native iOS application, then why is code written in a forbidden language, say Language X - which runs inside a native iOS application using a compatibility or a translation layer - prohibited? Since letting a third party layer of software come between the platform and the developer ultimately results in sub-standard apps and hinders the enhancement and progress of the platform, shouldn't JavaScript be banned too?

Double standards...

Saturday, August 14, 2010

Results of various tests related to IQ, logic, memory, reasoning, speed, etc., that I've taken

This post is intended to summarize the results of all the cognition-related tests that I've taken recently, and to replace all the related posts that I've written previously (all of which have been deleted).

The following chart from the now-defunct MSN Encarta shows a standard distribution of IQ scores.

The following list summarizes my performance (screenshots here):
  1. (6-Jun-06) (Result: "129")
  2. Super IQ Test (by Tickle) (10-Jan-07) (Result: Detailed result currently not accessible on the website. The only information available on record is that I was termed a Creative Theorist)
  3. IQ Test (by the now-defunct UberMens International) (11-Aug-08) (Result: "154")
  4. IQ Test (25-Dec-08) (Result: "137")
  5. Real Superhumans (memory test) (24-May-09) (Result: Scored above both "Average People" and "Super Humans")
  6. Real Superhumans (memory test) (24-May-09) (Result: Identical to the first time)
  7. IQ test (13-Jun-09) (Result: "over 140")
  8. IQ Test (13-Jun-09) (Result: "140")
  9. Ultimate IQ Test (21-Jun-09) (Result: "123")
  10. Ultimate IQ Test (based on Mensa IQ Workout) (21-Jun-09) (Result: "148")
  11. What's Your IQ? (11-Aug-10) (Result: "146")
  12. How's Your IQ? (11-Aug-10) (Result: "80", on a 0-100 scale)
  13. Kaplan 20 Minute Workout (GMAT) (17-Sep-10) (Result: "16/16")

Friday, August 06, 2010

Why I'm sad at the death of Google Wave and what I believe about the future of its innovations

I remember those weeks of last year when a buzz started around Google Wave - with the Sci/Tech section of Google News filled almost each day with stories about Wave. For a few days, I kept myself from reading about it, so that I could focus on work, in the hope that I would look at Wave in detail over some weekend.

I first got to know Wave when I watched a few videos of Google's I/O developer conference. I clearly remember clapping many times as the video unveiled feature-after-feature of Wave.

I was fascinated. Some of the things I had been thinking about for a few years were the design principles of Wave - no forwarding of messages (it unnecessarily duplicates messages, wasting tons of storage space) or attachments (a single copy that's pulled-on-demand). More lovely features - comment/reply right next to the content that your comment/reply applies to (wow!), playback a conversation (think before you Wave), a context sensitive spell-checker/correcter (this is how you do spell-correction), permissions built right into the design, drag-and-drop support, real-time updates, Robots (what a bloody fantastic idea), and Gadgets (another '!' idea).

Wave was billed as the future of electronic communication - the way modern communication ought to be done. It blended the myriad of communication methods we use today, and there was no reason Wave couldn't absorb yet more of the discrete ways of contemporary communication (voice, video, etc.). Wave has potentially many applications - comments (on blogs, websites), collaborative communication, document creation, social networking (imagine Facebook's Messages or Wall features built using Wave), wikis, and so many other things we can't even imagine right now.

I saw tremendous power in Wave, and no matter how unfriendly it might have appeared initially, Wave is not a bad idea. Bloggers and so-called "tech journalists" might have invested thousands of man-hours on initially gushing about and eventually bashing Wave, what they didn't grasp (and still haven't grasped) is that Wave is a fundamentally novel idea. The current UI of Wave is just one of the possible implementations of its features and underlying design principles, and it's perfectly possible to redesign the UI while preserving Wave functionally, if the need be.

However, fact of the matter is that Google is discontinuing Wave. I personally believe that it's an impatient and a premature move by Google. A bad move. Discontinuation of Wave only months after its public rollout and only over a year after its unveiling reveals potentially dangerous impatience and a severe lack of persistence, more so for a product billed as the future of communication (with Xbox, Microsoft has demonstrated that persistence pays, and Bing has begun paying off as well).

I very strongly believe in the fundamental rightness of Wave's design principles and its many features, and I predict that these will be back soon, either from Google or from others.

Tuesday, August 03, 2010

Post the committal of a crime, who deserves 'justice', the victim or the victim's family?

I was outraged Monday morning when I read a news story in The Tribune which said that murder charges against three Punjabi men (and the death sentence previously awarded) had been dropped by a UAE court, as a compromise had been reached between the families of the convicts and the victim's family.

WTF, I thought!

A poor guy was killed by the three men, and just because the victim's family has agreed to accept so-called blood money from the families of the murderers (arranged by some apparently prominent Punjabi in the UAE), is the dropping of murder charges by the Sharjah court even legal, let alone morally justifiable?

I - a patriotic and proud Punjabi myself - strongly believe that they're completely unjustified.

By punishing the killers, who are we trying to give justice to? The victim, of course. It's he who lost his life. It's he who is no longer alive. And just because his family is poor and ready to accept money for their son's life, does it legally justify the dropped charges/ punishment? The court knows that the three guys committed the murder. And despite this, its dropping of charges seems to be equivalent to saying - Ok. We know you are killers, and we know that killers deserve the strictest of punishments. But so what if that guy died? We're not punishing you-the-murderers, because as long as the dead guy's family is fine with this monetary compromise you guys have reached, we're fine too! After all, if the families of victims of murder, rape, theft, etc., are able to reach a "compromise" with the convict, there really isn't any need for the court to intervene. There is no such thing as an "absolute law". This court isn't here to clean the society. It's here only to handle those situations in which the poor victim has no family. Got it?

Really, it's sick! A crime and a criminal is over and above the family. How can a court close this loop without giving justice to the victim?

UPDATE [APR'12]: This video about the murder of Indian fishermen by Italian marines echoes some of the points I've mentioned above.

Monday, August 02, 2010

Infidelity and the cable TV

The other day, I observed an 18 year old guy watching TV. He was primarily watching a Hindi movie (which he had seen N number of times previously), but that doesn't mean that he wasn't also watching other stuff as well. What that means is that as soon an uninteresting scene or an (obviously boring) song would come, in a fraction of a second he would start flipping through channels and settle at another channel which looked interesting-enough to serve him for a few minutes (till the movie got interesting again).

Every few minutes he would switch back to the Hindi movie to check if it got interesting again, and if not, he would flip channels again and temporarily settle at something worthy. After about 30 minutes of this drama, he stumbled upon a channel which was playing a Hindi movie which he liked more than the one he was watching. Without a second thought, he started watching this one, ditching the previous one. But that didn't mean that the infidelity with the movie ended. The flip-channels-whenever-sucks drama continued.

I believe that the above is a typical case with many (or most) of us. Does this practice promote infidelity (Uncyclopedia; Urban Dictionary; Wikipedia) in the real life? Does the ease with which channels can be switched decrease one's fidelity? Does the multitude of television channels available on contemporary cable services program individuals to be disloyal and opportunistic, in relationships as well? Conversely, do people who practice infidelity flip channels more often? I don't know the answer, but I feel this might be true, or there might be a connection (subtle, at least).

Translation suggestion bar in Google Chrome should add a 'Wrong suggestion!' button for feedback

So that a user can kind of indicate to the translation suggestion feature that the suggestion you've provided is invalid, because the language that you've detected in this page simply isn't present!

An example is the image below, where there's no Malay in the webpage. Instead of clicking 'Nope' (which can have many meanings, as far as feedback is concerned), I would rather click 'Wrong suggestion!' button to make the system learn, or to merely build a log of pages where the feature malfunctions.

When I used the work feedback in this post, I refer to the following classic feedback model.

Ideal feedback model (source)

Recollecting my mistake in CASTLE MathQ'2001

I jointly won the CASTLE MathQ'2001, a state-level mathematics competition held in late 2001, in which 900 candidates appeared. While I was happy that I scored the highest marks, even if jointly, I had a somewhat strong regret at that time (whose intensity, and perhaps importance, has faded with time).

I remember that on the test day, when the allotted time was just about to finish, I had reached at the last question of the question paper. I was probably at a back bench, so when the time-over alarm had sounded and the test administrator had started collecting the answer-sheets, I still had a few seconds left before my sheet was taken. I knew that I had moments left, and aggressive as I believe I generally am, it was my intent to extract even the last possible drops of marks from this competition, in order to maximize my position. I was in battle-fighting mode that day, I remember clearly.

My brain wasn't working as well at that moment as it had been working all through the test, and I concluded that in the few seconds left, I wouldn't be able to solve the last question to reach at an indisputable answer. What to do? Guesstimate? Or just leave? 1 mark for a correct answer, a negative 0.25 for an incorrect answer, and no marks for a blank.

Last page of my actual MathQ'2001 question paper

My winning certificate :)

Not sure why, but right before my answer-sheet was taken, I just marked the most-likely-looking option as the answer.

And it turned out to be incorrect. I lost an extra 0.25 mark that I wouldn't otherwise have lost. A narrow win became a joint-win. The first prize - a Vintron Biz123 desktop computer - changed into a joint prize. And I regretted this a lot (at that time at least).