lies,damned lies and statistics Essay

August 21, 2017 Communication

“Lies. damned prevarications. and statistics” is a phrase depicting the persuasive power of Numberss. peculiarly the usage of statistics to bolster weak statements. It is besides sometimes conversationally used to doubt statistics used to turn out an opponent’s point. The term was popularised in the United States by Mark Twain ( among others ) . who attributed it to the 19th-century British Prime Minister Benjamin Disraeli ( 1804–1881 ) : “There are three sorts of prevarications: prevarications. damned prevarications. and statistics. ” This Line stresses on the fact that common mistakes. both knowing and unwilled. associated with the reading of statistics. and how these mistakes can take to inaccurate decisions. It is a phrase attributed to the power associated with figures. The phrase is normally used to doubt statistics given to back up authorities places.

2. The political orientation behind this phrase
Numbers and expressions are supposed to stand for “objective scientific data” you can non deny and examined by intelligent and experient experts. Now the complete prevaricator wants his counterfeits to look undeniably “scientific” . so why non utilize the thaumaturgy of Numberss that the not-so-math-literate multitudes could ne’er deny? They say that statistics don’t prevarication. and while that may be true. prevaricators do utilize statistics. So it is with much that you read and hear. Averages and relationships and tendencies and graphs are non ever what they seem. There may be more in them than meets the oculus. and there may be a good trade less.

The secret linguistic communication of statistics. so appealing in a factminded civilization. is employed to sensationalize. blow up. confuse. and oversimplify. Statistical methods and statistical footings are necessary in describing the mass informations of societal and economic tendencies. concern conditions. ‘opinion’ polls. the nose count. But without authors who use the words with honestness and apprehension and readers who know what they mean. the consequence can merely be semantic bunk. 3. Use of this phrase in assorted topographic points: –

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

In advertisement and Politicss and other signifiers of propaganda:
This phrase covers all cases of Artistic License – Statistics where statistics are used to deceive people as to the truth. The job is. people do non pay attending to the context. merely the Numberss. For illustration. the statement “Brand X is 84 % fat-free” sounds good until you realize that this means the nutrient merchandise is 16 % fat by weight. Besides. “fastest growing” could intend that at that place used to be one client and so there were five more. doing a five-hundred per centum addition. You should besides detect Absolute Comparatives: it’s fastest turning. but specifically compared to when/what? The whole concern of throwing per centums at people in advertisement. political relations and other signifiers of propaganda is about destined for this sort of maltreatment. Relative steps are more likely to be understood accurately. and therefore are less likely to be used in advertisement. In insouciant links between things which are non related:

The fake utilizations of statistics are intended to connote a causal nexus between two elements when they are non linked. the nexus is questionable. or the nexus is opposite to what is implied. A beautiful illustration? “Coca-Cola causes drowning” . By looking at statistics on submerging and Coca-Cola gross revenues. you can see a nexus — more people go swimming on hot yearss. and more people buy Coke on hot yearss. Likewise. birth rates per caput of population are higher in countries where there are more storks — because birth rates are ever higher in rural countries. which is where one finds the Delivery Stork.

Correlation does non equal causing ; if it does. so we might besides reason that planetary heating is caused by a diminution in pirate population and that 100 % of Homo Sapiens who consume dihydrogen monoxide will discontinue critical maps and decompose. Besides be cognizant of the Law of Very Large Numbers. Any fraction of a really big figure is likely to be a big figure. no affair how little the fraction is. It is estimated that 2. 135. 000 Americans have used cocaine ( including cleft ) in the past month. But that’s merely 0. 7 % of the population! So. is this a batch of people. or non? In doing things more singular than they truly are:

You should besides be on the sentinel for the related consequence where things are made more singular than they truly are. The odds that any given ticket will win a raffle may be really little. but it is certain that one will be a victor. You’d notice being dealt a royal flower in spades at fire hook. but the odds of it go oning are precisely the same as those for being dealt any other manus of five specified cards. You can turn out anything utilizing statistics:

Statisticss are like surveies: who made them and who paid them matters a batch. Want to “prove” that picture games cause force? Get a group of scientists that are already savvy to this and don’t head the deficiency of moralss. Have them pull from a really little pool of trial topics that are known to expose violent behaviour. Mental infirmaries. prisons. schools for kids with behaviour upsets. what have you. Make some generic trials that are guaranteed to demo up positive. come up with Numberss. and presto. instant headline. “Recent trial shows 77 % of topics become more violent after playing Mortal Kombat. ” Most people won’t bother with reading the article the whole manner through and will merely look at the headline. This works with anything from amusing books. and stone to watching Brokeback Mountain or vote for specific parties. fundamentally anything.

4. There are a batch of illustrations which attest to this phrase and turn out this statement right There are illustrations from about every facet where statistics has been used to reason a assortment of false or uncomplete decisions. Some of them are: – 1. On a historical Note: –

Something of a historical corruption: During World War II. the Royal Air Force wanted to add more armour to their planes. but because of weight bounds they needed to cognize which topographic points needed the armour most. So. they examined the planes after they came back and counted how frequently bullet holes were found in certain areas… and so placed armour in topographic points that showed the fewest slug holes. This is because. they assumed. that any topographic point that did hold slug holes was a topographic point that planes could be hit and still wing. Helped by the fact: No plane that of all time came back had holes where the gas armored combat vehicle was. Because planes whose armored combat vehicle was hit would detonate and non come back. 2. Pathetic Decisions: –

It’s a spot like the “statistics” on shark shows. “You are more likely to decease on the lavatory than be eaten by a shark. ” When you compare how much clip you spend around sharks versus how much clip you spend around lavatories … truly. the lavatory has clip to be after out its move in progress. Same trade with most accidents happening in the place. Sing that you spend the bulk of your clip in your place. this should come as no surprise to anyone.

The same for the illustration above about most vehicular accidents happening near the place ( some say “within 25 stat mis from your home” ) . This is because most people do most of their drive near their places. non that the place or the encompassing country is more unsafe than countries distant from the place.

3. For Doctors: –
Nine out of Ten Doctors Agree that the phrase “Nine out of Ten Doctors Agree” has been practically a stock phrase in advertisement since the early twentieth century. “Nine out of 10 tooth doctors recommend Trident for their patients who chew gum. ” The 10th tooth doctor was repetitive that his patients ne’er chew gum at all. but surprisingly. Trident didn’t want you to cognize about that.

One interesting instance happened in Portugal. where two ads were being broadcasted on national Television during the same period ( and sometimes even in the same commercial interruption ) claiming. severally. that ‘90 % of tooth doctors use toothpaste X’ and ‘8 out of 10 tooth doctors recommend toothpaste Y to their family’ . Together. if you stop to believe about it. they imply something is non rather right about those professionals’ concern over their ain family… Or that an atrocious batch of tooth doctors are single orphans. hence can’t recommend it to a household they haven’t got. 4. Ad Campaign

In Montreal. there was an ad run run by a gum company whose gum came in unit of ammunition forms alternatively of the usual square forms. The ad said. “100 % of people who chew square gum dice. ” 5. Casinos: – Many casinos like to publicize their slot machines with lines like “Up To 99 % Payout! ” to do it sound like the participant has a good opportunity to win. First. “up to” means the payout could be 1 % for all you know ( although Torahs normally set a lower limit ) . Second. even a 99 % payout means that for every $ 100 you put in the machine. on norm. you’ll acquire $ 99 back. i. e. you still lose. That “99 % payout” is besides an norm that is based on something like one million pulls ( dramas ) on the machine. If you play 100 times in one slot machine. you’re non acquiring a representative sample of that norm.

6. Plans On Television
Plans on Animal Planet are fond of mentioning how Americans spend more money yearly on cat or Canis familiaris nutrient than on babe nutrient. This is depicted as grounds that Americans pamper their pets like babes. but overlooks several facts: that pets eat favored nutrient for their full lives. whereas babes merely eat babes nutrient for about a twelvemonth and a half. and that many households have more than one pet at a clip. but comparatively few have more than one kid of an age to eat babe nutrient at the same clip. 7. Car Insurance

Ever inquire how all auto insurance companies manage to to publicize that “people who switch from to salvage an norm of “ ? It’s because the sample population “people who switch” is about wholly composed of “people who are traveling to salvage a large ball of money making so” . or else why would they trouble oneself to exchange? Since no record is kept of the per centum of people who would non salvage any money and hence don’t switch. the cited statistic has about no significance. 8. A Paradox: –

Simpson’s Paradox is when information shows one tendency. but spliting it into classs shows the opposite tendency. In the illustration above. infirmary 1 has a higher decease rate. but if the patients are split into classs based on badness of hurt. it has a lower decease rate in each class. The same goes with good physicians and bad physicians. as told in the book Super Freakonomics. Good physicians are by and large given tougher causes while bad physicians are given easier instances.

However. if you look at decease rates you see that some physicians have higher decease rates. but these are normally the good physicians. Patients with serious instances are more likely to decease. so good physicians lose a batch of their patients than. state the physician who cures hiccoughs. The lesson is that you can be reasonably certain that the physician you receive at a infirmary is competent plenty to be assigned to you.

5. How to guarantee that we do non lie with the aid of statistics There are a figure of issues which need to be addressed in order to guarantee that mistakes. both knowing and unwilled. associated with the reading of statistics. are minimized. 1. Sampling Biass

Response Bias: Inclination for people to over- or under-state the
truthNon-response: Peoples who complete studies are consistently different from those who fail to react. Accessibility/Pride. Representative Sample: One where all beginnings of prejudices have been removed. ( Literary Digest ) Questionnaire wording/Interviewer effects

Recall Bias: Inclination for one group to retrieve anterior exposure in retrospective surveies 2. Happy Average Arithmetical Mean: Evenly distributes the sum among persons. Can be unrepresentative when measurings are extremely skewed right. ( e. g. per capita income ) Median: Value spliting distribution into two equal parts. 50th percentile. ( e. g. average household income ) Manner: Most often observed result ( seldom reported with numeral informations ) 3. Small Figures Not There

Small samples: Calculators with big standard mistakes. can supply apparently really strong effects Low incidence rates: Necessitate really big samples for meaningful estimations of low frequence events Significance levels/margins of mistake: Measures of the strength and preciseness of illation Scopes: Report ranges or standard divergences along with agencies ( e. g. “normal” ranges ) Inferring among persons versus populations

Clearly label chart axes
4. Much Ado About Nothing
Probable Mistake: Appraisal mistake with chance 0. 5. If calculator is about normal. PE is about 0. 675 standard mistakes. ( Old school ) Margin of Mistake: Appraisal mistake with chance 0. 95. If calculator is about normal. PE is about 2 standard mistakes Clinical ( practical ) significance: In really big samples an consequence may be important statistically. but non in a practical sense. Report assurance intervals every bit good as P-values.

5. Attention-getting Graphs
Choice of scopes on graphs can hold immense impact on reading ( e. g. per centum alteration ) Choice of proportion of y-axis to x-axis can falsify every bit good ( really easily to make with modern package ) Can besides distort saloon charts by holding them start at positive values and/or paring below an unreal baseline to 0

6. 1-D Pictures
Bar Charts and Pictorial Graphs should hold countries relative to values ( merely make comparings in one dimension ) 7. Semiattached Figure Target Population: Group we want to do illation sing Study Population: Group or points that experiment or study is conducted on When comparative surveies are conducted among merchandises. interventions. or groups ; what is the comparison merchandise. intervention. or group? Control for all other possible hazard factors when analyzing effects of factors 8. Causal Relationships

Correlation does non connote causing
Elementss of causal relationships
a. Association between Y and X
B. Clear clip telling ( X precedes Y )
c. Removal of alternate accounts ( commanding for other factors ) d. Dose-Response ( when possible )



Decision

In the terminal. statistics are non prevarications and statistics don’t prevarication: people lie about the statistic itself or how it is interpreted. Some don’t prevarication. they are merely nescient. as are most members of the populace in footings of statistical reading. See Logical Fallacies and Critical Research Failure. Put another manner. by baseball announcer Vin Scully:

“People usage statistics the manner a rummy uses a lamp station — for support. non light. ”

x

Hi!
I'm Amanda

Would you like to get a custom essay? How about receiving a customized one?

Check it out