The Free Dictionary  
mailing list For webmasters
Welcome Guest Forum Search | Active Topics | Members

±15.4% Options
Aventador LP700-4
Posted: Wednesday, June 28, 2017 6:31:36 AM
Rank: Member

Joined: 2/13/2017
Posts: 49
Neurons: 4,372
What does the number " ±15.4% " refer to in this context?

According to the crash stats report, “brake related problems as critical reasons accounted for about 22 percent (±15.4%)” of crashes where vehicle failure was the cited cause of the crash.


Source:
http://www.lowmanlawfirm.com/blog/bid/63142/top-5-mechanical-failures-that-cause-car-accidents

If the result of the report fluctuates between 15 to 16 percent out of 22 percent, then the result can be so inaccurate and doesn't worth announcing!

It doesn't make sense!
pjharvey
Posted: Wednesday, June 28, 2017 6:57:22 AM
Rank: Advanced Member

Joined: 4/13/2012
Posts: 754
Neurons: 42,851
I really don't understand: as a tolerance to number 22, plus/minus 15.4 does not make sense (a range from 6.6 to 37.4 is absurd!).
Does anyone understand?

Unless it is a percentage of a percentage, that is, 22 plus/minus about 3.4, which would make a range from about 18.6 to about 25.4.
Orson Burleigh
Posted: Wednesday, June 28, 2017 7:04:57 AM

Rank: Advanced Member

Joined: 10/12/2011
Posts: 137
Neurons: 47,648
Location: Annapolis, Maryland, United States
Aventador LP700-4 wrote:
What does the number " ±15.4% " refer to in this context?

According to the crash stats report, “brake related problems as critical reasons accounted for about 22 percent (±15.4%)” of crashes where vehicle failure was the cited cause of the crash.


Source:
http://www.lowmanlawfirm.com/blog/bid/63142/top-5-mechanical-failures-that-cause-car-accidents

If the result of the report fluctuates between 15 to 16 percent out of 22 percent, then the result can be so inaccurate and doesn't worth announcing!

It doesn't make sense!


Applause You are right - announcing an estimate of 22% with a margin of error of (±15.4%)” is hardly worth the breath used to make the announcement.

The law firm which provided the statistic in its blog is apparently engaging in the practice of ambulance chasing.
thar
Posted: Wednesday, June 28, 2017 7:06:49 AM

Rank: Advanced Member

Joined: 7/8/2010
Posts: 15,290
Neurons: 60,813
It is a law firm.
There are lies, damn lies, statistics and compensation lawyers!

I agree, it should be a range of the percentages (you would never give a percentage of a percentage, surely!)

And what is their statistical method, anyway?

If they have a certain dataset, they can interrogate it - how many of these crashes (which are due to vehicle failure) list brake related problems as a critical reason? x out of y. That is a straight ratio. 22/100.

If you have lots of different datasets, why haven't you combined them? Because they have different criteria? Why can't they be compiled? Are crashes in Fords more likely to be due to brake failure? Or crashes in France? Or crashes on the motorway?

This is not process that produces a range of data values. It seems to be a yes/no question, to me.

Deliberate obfuscation by a law firm?


TMe
Posted: Wednesday, June 28, 2017 10:58:55 AM

Rank: Advanced Member

Joined: 1/12/2017
Posts: 356
Neurons: 2,279
Means 15.4% of 22 either way.
It can be 22-(15.4% of 22) i.e. ).825= 21.175

OR

It can be 22+(15.4% of 22) i.e ).825= 22.825

That mean margin runs between 22.825 and 21.175.IMO

According to the crash stats report, “brake related problems as critical reasons accounted for about 22.825 and 21.172 percent of crashes where vehicle failure was the cited cause of the crash.



I am subject to correction.


I am a layman.
Харбин Хэйлунцзян 1
Posted: Wednesday, June 28, 2017 12:38:28 PM

Rank: Advanced Member

Joined: 10/13/2015
Posts: 701
Neurons: 182,268
Location: Dzerzhinskiy, Moskovskaya, Russia
thar wrote:
I agree, it should be a range of the percentages (you would never give a percentage of a percentage, surely!)

But it looks like it is a percentage of a percentage. At least this was my impression when I read the original:

http://www.docslides.com/test/a-brief-statistical-summary

In addition to 22% ± 15.4% they have 3% ± 3.3% in table 3:





აბა ყვავებს ვინ დაიჭერს, კარგო? გალიაში ბულბულები ზიან.
Kunstniete
Posted: Thursday, June 29, 2017 1:40:59 AM

Rank: Member

Joined: 1/25/2017
Posts: 902
Neurons: 104,088
Location: Berlin, Berlin, Germany
Descriptions of %-values seem to confuse people more than you'd think. Just the other day I saw an article about the need for economists which reads: "Almost every fifth job advertisement (4,5%) is a call to economists". Clearly, there is a huge difference between every fifth & 4,5%, but maybe the author was drunk or unconscious while writing the article Whistle

As for the current topic: I agree with thar, a percentage of a percentage is confusing & senseless.

Also, there a more elegant & subtle ways to manipulate your data ;)

The value of choice is not in the size of the action but in its effect.
Drag0nspeaker
Posted: Thursday, June 29, 2017 5:05:16 PM

Rank: Advanced Member

Joined: 9/12/2011
Posts: 26,242
Neurons: 139,836
Location: Livingston, Scotland, United Kingdom
Kunstniete wrote:
Also, there a more elegant & subtle ways to manipulate your data ;)

Don't tell everyone Shhh - the newspapers will start doing it! Whistle

Wyrd bið ful aræd - bull!
Users browsing this topic
Guest


Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

Main Forum RSS : RSS
Forum Terms and Guidelines. Copyright © 2008-2017 Farlex, Inc. All rights reserved.