- HOME
- INTRO TO THE FORUM
- BADLY WRITTEN, BADLY SPOKEN
- GETTING
TO KNOW ENGLISH - PREPARING FOR ENGLISH PROFICIENCY TESTS
- GOING DEEPER INTO ENGLISH
- YOU ASKED ME THIS QUESTION
- EDUCATION AND TEACHING FORUM
- ADVICE AND DISSENT
- MY MEDIA ENGLISH WATCH
- STUDENTS' SOUNDING BOARD
- LANGUAGE HUMOR AT ITS FINEST
- THE LOUNGE
- NOTABLE WORKS BY OUR VERY OWN
- ESSAYS BY JOSE CARILLO
- Looking Back to Easter Sunday’s Earthly and Celestial Foundations
- Our Need For Thinking National Leaders With The Gift Of Language
- The Heavy Price of Misplacing One’s Trust and Confidence
- The Three Basic Word-Positioning Principles For Emphasizing Ideas
- Use the Reflexive Form When The Verb’s Object is The Doer Itself
- Ferreting Out Truth And Falsehood From The Flood of Poll Propaganda
- ABOUT JOSE CARILLO
- READINGS ABOUT LANGUAGE
- TIME OUT FROM ENGLISH GRAMMAR
- NEWS AND COMMENTARY
- BOOKSHOP
- ARCHIVES
POUR OUT YOUR MIND IN ENGLISH
This page welcomes well-thought-out and well-considered postings in English about any subject under the sun. Postings that will be entertained are original essays, short stories, and verses in the English language written by the forum member himself or herself.
The maximum allowable posting is 1,200 words in 12 points in the standard 8 ½” x 11” typeset format; longer but worthwhile postings will be cut or condensed by the moderator to fit the maximum space allocation.
The objective of this page is to promote and encourage the felicitous use of English in expressing ideas, thoughts, and feelings. By posting material on this page, the member expressly agrees to have it critiqued by fellow members of Jose Carillo’s English Forum. Members making critiques, on the other hand, are enjoined to make their critiques constructive, thoughtful, and helpful.
Click here to go to the board (requires registration to post)
And then there are statistics
By Renzphotography
And then there are statistics!
In great disgust, one American politician quipped that “there are lies, there are bigger lies, and then there are statistics.”
To my mind, nothing is more deliberate and insulting than to use math or science to justify lies. It is simply a blatant exploitation of people’s lack of knowledge—in short, their ignorance—about a certain field of study. While there are professionals who can be consulted for alternative opinions on certain scientific and mathematical questions, the one peculiar problem inherent to statistics lies in the way it is implemented.
I am trained in statistics and I thought that big-ticket survey outfits were very scientific and methodical in implementing statistical studies. However, I was rudely awakened when, upon close scrutiny—through an interview with experts from one of these outfits years ago—I realized that the entire rule book for conducting proper statistical surveys is simply thrown out of the window.
The major flaws are in defining the scope of the survey and in the actual gathering of data. Just as in any scientific and mathematical endeavor, it is important to properly define the scope and limitations of any statistical survey. Since statistics is the branch of mathematics that describes a population based on random surveys on samples, then it makes sense to take enough and correct samples or respondents within the appropriate scope of the study.
Let us put this in the proper context. Suppose we conduct a survey that aims to describe a group called “Five-block-village,” the population of which is separated by five blocks. Then it makes sense to take samples or respondents from all five blocks.
If all samples are taken from just one block, say the “Alpha” block, then the results must be called the survey results from “Alpha” block and not from “Five-block-village.” This is because the results were drawn exclusively from just one block.
In fact, if the “Five-block-village” has a population of 1,000 and if the target respondent size is 100, then it makes sense to get 20 respondents from each block.
Now, if there is a total of 100 people in block “Alpha,” 50 in “Bravo,” 400 in “Charlie,” 200 in “Delta,” and 250 in “Echo,” then the best way to distribute the number of target respondents is to follow the proportion of the distribution of the population. Therefore, it makes sense to take 10 respondents from block “Alpha,” 5 from “Bravo,” 40 from “Charlie,” 20 from “Delta,” and 25 from “Echo.”
In television rating surveys, pollster firms install an expensive device on TV sets that monitors the TV stations viewed at particular time slots. I know for a fact that these devices are installed in upscale and middle-class households, but for safety reasons, very reluctantly—if ever—on poor households. So, how do these polling firms survey and describe the behavior of the viewers in the poor households? You guessed it, they use their magic wands!
Another problem here is that most of the machines for such poll surveys are installed in Metro Manila households. So, do we conclude that Metro Manilans have the same viewing preferences as those in other parts of the country?
Now, what if I tell you that the same survey firms are behind the poll ratings of the presidential candidates? And what if I tell you that the same sampling flaws are conducted when describing the nation’s presidential preferences—that is, when the respondents are mostly from the upper segments of Metro Manila society?
Not all surveys are flawed, however. The one survey that I admire for accuracy and depth is the one prepared by the National Statistics Office on the population every five to ten years. It is a painstaking process where statisticians walk from house to house to gather data. It takes one whole year to prepare and the project is so costly that it is conducted several years in between.
So, whenever we see another presidential survey report, perhaps we should pause and think twice. If the survey company can update the results almost daily, then there has got to be something wrong.
Click to read comments or post a comment
View the complete list of postings in this section
