Reporter WTF



  • The ecnomy is bad, it only adds insult to injury when those reporting on it show signs they don't understand basic third grade mathematics (no wonder why nobody saw it coming):

    http://www.baltimoresun.com/business/realestate/bal-home-prices0512,0,3625468.story

    Read the first line under the headline, then read the first sentance of the first paragraph.

     

    The thing is that her job is simply to paraphrase someone else's work if it isn't a field that she can understand. Maybe she should have just copy and pasted or did some investigative reporting into what a median is.

    Or, maybe she is wikipedia material, at least that way someone can correct her mistakes.





  •  In the interest of not confusing stupid people any further, mathematicians got together and decided to use average and median interchangably.  They also selected a new term to mean the same thing "median" used to mean: "imatard."  Be sure to use it often.



  • The sentence under the headline says average, and the rest of the article states more specifically the median, as opposed to the more commonly assumed mean?



  • @ComputerForumUser said:

    The sentence under the headline says average, and the rest of the article states more specifically the median, as opposed to the more commonly assumed mean?

    Wait, "mean" isn't just an abbreviation for "median"?!



  • @ComputerForumUser said:

    The sentence under the headline says average, and the rest of the article states more specifically the median, as opposed to the more commonly assumed mean?

     

    Actually, the "average" refers to the percentage drop of the "median" price, and the comparison is of the fall in that area compared to the national average fall.

    By fall I mean drop, not the season, of course.

     



  • @BillC said:

    By fall I mean drop, not the season, of course.
     

    You refer to the tiny portion of liquid, but not additional flavoring, of navigational direction?



  • @dhromed said:

    @BillC said:

    By fall I mean drop, not the season, of course.
     

    You refer to the tiny portion of liquid, but not additional flavoring, of navigational direction?

    He hasn't yet, but he will by autumn.



  • @Pol said:

    http://www.baltimoresun.com/business/realestate/bal-home-prices0512,0,3625468.story

    ...for the hard of copying and pasting

     

    +1 Informative



  • In my experience, "average" can refer to arithmetic mean, median or RMS. Obviously in the context of housing prices, it means median.

    However, this is my problem with the article:

    The Baltimore area's median price, which had been $270,500 in the first quarter of 2008, declined less sharply than the national median, which plummeted 13.8 percent to $169,000, NAR said in its quarterly report.

    Do these numbers make sense? Is $270500 minus 13.8% anywhere near $169000?




  • Is it permissible to be mean if one is above average?

     



  • This entire thread is TRWTF.

     OP: Mean, median and mode are all types of average or 'central tendency' therefore I see no WTF in the article.

     tOmcOlins: The numbers do make sense. 13.8% and $169,000 are refering to the national values while 9% and $270,500 are specific to the Baltimore area. As it says quite clearly in the line you quoted...



  • @tOmcOlins said:

    The Baltimore area's median price, which had been $270,500 in the first quarter of 2008, declined less sharply than the national median, which plummeted 13.8 percent to $169,000, NAR said in its quarterly report.

    Do these numbers make sense? Is $270500 minus 13.8% anywhere near $169000?

    They make sense, but they aren't particularly useful together:

     

    LocationJuly 08NowDecline
    Baltimore$270,500$245,8009%
    National?$169,00013.8%

    The stuff in bold is what is contained in that sentence, which makes comparisons rather difficult, but the numbers presumably make sense (provided the national median really did fall 13.8%). But why is the Baltimore decline figure rounded down to 9 instead of the more accurate 9.1%?



  • Okay, that makes sense now.

    Although it was my fault for not reading it correctly, I'm not going to admit it and blame the author for writing it poorly.



  • @Vechni said:

    The thing is that her job is simply to paraphrase someone else's work

     I wonder why so many newspapers, including The Baltimore Sun, are going bankrupt?

    Oh, right, they hire typists, not reporters.



  •  1) a quantity, rating, or the like that represents or approximates an arithmetic mean

    You could say that the author isn't a math expert, perhaps not an english expert either. Now I can't read the word average and interpret it as median because that makes no sense to me... I guess some people can do that. However that isn't a defense when the dictionary disagrees with your writing, and you're a writer.

     

    My actual gripe is more along the lines that the author probably doesn't even know what a median is (the actual one, apparently), and is responsible for helping to keep the public informed. Wouldn't be suprised if she couldn't come up with the formula for an average and thinks that the mean price = going price.



  • @Mithious said:

    This entire thread is TRWTF.

     OP: Mean, median and mode are all types of average or 'central tendency' therefore I see no WTF in the article.

    TRWTF is that I used to work next to someone with a masters in statistics, who claimed she'd never heard such nonsense, and claimed every source I pulled to show her that it was true was wrong.  I gave up after showing her five different books on statistics (she never brought in any competing documentation; she had a degree, so her word apparently trumped any book brought by someone without a competing degree), as I happened across one of her old classmates.  He indicated that she'd claimed that to be nonsense to the professor the day it was covered in class - and that it was in their book, too.

    Granted, calling 'mode' an 'average' is one way certain unscrupulous people manage to lie with statistics, as sometimes the mode is substantially skewed from either the mean or median.  In fact, I think it's pretty much only likely to be close to the other averages if it's a bell curve distribution.  I know with exponential distributions, the mode is always more or less at the functional minimum.

    Also granted, the lack of competency she presented in her day job did not give any indication that she'd have had any real competency at her academic specialties.  It was apparently me giving the benefit of doubt to one who could not utilize it.  For what it's worth, this woman later made the claim that a survey population of 5 individuals from the same department of 8 people could be statistically significant across a company of 100k people, because there was a very low standard deviation in their response.  She presented this in a meeting of roughly 20 people, 19 of whom disagreed with her and the 5 members of her department in her survey...



  • @Vechni said:

    Now I can't read the word average and interpret it as median because that makes no sense to me...

    Hypothetical case for you. Consider the following data points (sorted) : 4, 4, 4, 4, 4, 4, 4, 5, 5, 6, 8, 14, 20, 150, 3011

    For this series, which number best represents the concept of 'average'? 216 (the mean), 5 (the median), or 4 (the mode)? If you're using the average to attempt to predict what number may be generated next (remember, I sorted the numbers, they weren't generated in that order. Theoretically speaking, anyway), I'd recommend using either 5 or 4, as the mean is greater than any number other than the maximum.

    No, I'm not going to let you use a trimmed mean value of 17.  That's cheating.  ;> (A trimmed mean value of 6 is, of course, cheating even more.)



  • There is in fact one WTF in the article, but amusingly it's not anything suggested previously in this thread.

    The 9% drop is described as being "in the first three months of the year", but a careful reading of the article will show that in fact the quoted statistics are all based on a year-on-year comparison between first quarter 2008 and first quarter 2009 (this is a good thing to do because it avoids seasonal fluctuations). The 9% drop is therefore over the entire previous year, not just the first three months of 2009.


  • Discourse touched me in a no-no place

     TRWTF is anyone expecting any sort of journalistic integrity, or any manner of correctness, or even anything that EVEN BEGINS TO MAKE SENSE from the Baltimore Sun.




  • @tgape said:

    Hypothetical case for you. Consider the following data points (sorted) : 4, 4, 4, 4, 4, 4, 4, 5, 5, 6, 8, 14, 20, 150, 3011

    For this series, which number best represents the concept of 'average'? 216 (the mean), 5 (the median), or 4 (the mode)?

     

    The question is: what's the generally accepted method of eliminating obvious anomalies such as 3011 from that set? In other words, what's the accepted definition of "anomalous"? (real question btw thxbye)



  • @dhromed said:

    @tgape said:

    Hypothetical case for you. Consider the following data points (sorted) : 4, 4, 4, 4, 4, 4, 4, 5, 5, 6, 8, 14, 20, 150, 3011

    For this series, which number best represents the concept of 'average'? 216 (the mean), 5 (the median), or 4 (the mode)?

     

    The question is: what's the generally accepted method of eliminating obvious anomalies such as 3011 from that set? In other words, what's the accepted definition of "anomalous"? (real question btw thxbye)

    Depends on what the data represents, and whether it's actually desirable to eliminate such data. http://en.wikipedia.org/wiki/Outlier probably describes what you're after.


  • @PJH said:

    Depends on what the data represents, and whether it's actually desirable to eliminate such data. http://en.wikipedia.org/wiki/Outlier probably describes what you're after.
     

    Thankz. I'm having an uprising of interest in statistical math. Also wondering why the Bayesian method differed in outcome from the classical method by a huge 5%, when applied to the boys & girls & pants example in the Bayes wiki article. I'm probably doin it rong.

    EDIT

    Oh, wait, ofcourse. I misdefined the total population size. it's 80, not 100. Discrepancy gone. So why is the Bayesian method helpful again?



  • @Weng said:

     TRWTF is anyone expecting any sort of journalistic integrity, or any manner of correctness, or even anything that EVEN BEGINS TO MAKE SENSE from the Baltimore Sun.

     

    In my experiance, the Sun reprints press releases. Especially in regard to Baltimore City government press releases, often word for word. So if the press release is correct and makes sense, the Sun would be both correct and sense-making. Especially in regard to Baltimore City government press releases, this rarely happens.

    Blaming the Sun is like blaming your TV for the poor acting and bad dialogue in a show. I can only assume you are confusing The Sun with a newspaper.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.