Log in

No account? Create an account
Smiling Headshot

Statistics, Sources and Damn Lies

Posted on 2009.08.14 at 16:42
So, a coworker (who I actually like) sent out the following link today:

Statistics Show Social Media Is Bigger Than You Think

The point of the article is to use startling facts to get people who aren't involved in the Social Networking space to realize its potential. I suspect it is half 'ra-ra go Social Networking' and half 'if you are a 50+ y/o CEO who doesn't know what Twittering is, this is why you should pay attention.' ETA: This is actually a blog maintained by someone who just released a book on this space

However, because I'm a bad person, I distrust such 'facts' off the bat. While at least article lists sources, many of the sources are summary links of other sources.

1. By 2010 Gen Y will outnumber Baby Boomers. 96% of them have joined a social network Source, which is a single line summary of a study

Okay, so the source states the following:
96% of online teens/tweens report ever having used any type of social networking technology including IM/chat, text messaging and email. And yes, Gen Y outnumbers the baby boomers, just like the boomers outnumbered their parents. A similar statement could have been made in the year 1950: Baby Boomers outnumber their parents, and 96% of them watch TV (as defined by viewing at least 1 hour in their life).

The real crime is equating 'having used any type... including... email' (emphasis mine) as 'joined a social network'.

2. Social Media has overtaken porn as the #1 activity on the Web Huffington Post Source, which is really just a link to a Reuters article

Strike 1: Not actually referencing the source.
Strike 2: Stating as fact, something which was stated as being from a study, which in reality was an authorial comment based on a limited research and correlation as causation.
Strike 3: The only actual comment was that searching for porn [whatever that means] decreased from being 20% of the searches to 10%. Now I can come up with a lot of reasons for that, but most notably most porn is easier to find than in the 'old days' where it was all behind password protected sites.
Laughter value: Quote from the interview with the Author: But Tancer said the speed at which information spread on the Internet had meant in some cases it was consumers generating the story and the media is last to record it -- or fact-check it.

"With the explosion of this type of false information on the Internet I think we will see someone come forward and develop a new type of software that can filter for the most accurate information," he said.

"Maybe accuracy is the next thing we will all search for."

3. 1 out of 8 couples married in the U.S. last year met via social media (Source is a link to a single line verbatim referencing a McKinsey study. Googling around got me to here which has a link to a protected doc called: Advertising Age – McKinsey Study Predicts Continuing Decline in TV Selling Power
McKinsey & Co. is telling a host of major marketers that by 2010, traditional TV advertising will be one-third as effective as it was in 1990.

So, no idea on the actual context surrounding that statement or what 'met' means nor what 'social media' means. Because, as per above, that could clearly mean that they used any personals ad online rather than an actual social network. But whatever, big news: People use online dating!

4. Years to Reach 50 millions Users: Radio (38 Years), TV (13 Years), Internet (4 Years), iPod (3 Years)…Facebook added 100 million users in less than 9 months…iPhone applications hit 1 billion in 9 months.
(several sources, the first of which and covers the first 3 just lists the stats again)
1) Radio audience was measure in terms of households, not individual users. Cost to Entry initial: Very high.
2) TV, see above.
3) I have no idea what they are using for audience for internet, but internet users were able to largely use existing infrastructure and many first connected through work, which was not the case for TV/Radio.
4) Selling 50 million iPods != 50 Million users (rebuying from broken, having multiple iPods for different uses, etc.)
5) Facebook you could have many accounts, and many people do for taking advantage of the games. When I was doing those heavily several people sent me their doubles. Easily 10% of my 'friends' were doubled accounts.
6) WOW MISLEADING iPhone. 1 billion downloads is sooo not anything related to # of users (which is not audience and ugh).


5. If Facebook were a country it would be the world’s 4th largest between the United States and Indonesia

See above discussion of multiple accounts for a personal user and lets not even get into the company other pages and people re-creating under new accounts to hide from other friends.

6. Yet, some sources say China’s QZone is larger with over 300 million using their services (Facebook’s ban in China plays into this) Source

No clue as the source is self reported and subject to scrutiny. But, the source says 200 million. What is 100 million, plus or minus between friends, eh?

7. comScore indicates that Russia has the most engage social media audience with visitors spending 6.6 hours and viewing 1,307 pages per visitor per month – Vkontakte.ru is the #1 social network

Unlinked source, but my findings actually show that while Russia is the fasting GROWING market, it ranked last with penetration for those numbers. Linked Source

8. 2009 US Department of Education study revealed that on average, online students out performed those receiving face-to-face instruction (Another source that links to the actual damn study. PDF Study WTF LINK TO PRIMARY SOURCE).

Fun part about this fact: This contrast falls just short of statistical significance (p < .06) when the five K-12 contrasts are removed from the analysis.

Hint, if it is barely statistically significant, perhaps you shouldn't trumpet this as undisputed fact, huh?

9. 1 in 6 higher education students are enrolled in online curriculum
Source is attempting to relocation. How about: Your ass.

10. % of companies using LinkedIn as a primary tool to find employees….80%
What they mean is that 80% of 80 (!!!) self reporting companies who belong to Jobvite, which self-professes to be focused on next generation recruitment solutions, reported on planning to use social networks at all during a search for candidates. What the survey actually revealed was that the primary tool was: Employee Referrals.

11. The fastest growing segment on Facebook is 55-65 year-old females
Old news, and many have already commented that this due largely to the other segments being well penetrated. OTOH, my mom falls into that segment and just joined Facebook.

12. Ashton Kutcher and Ellen Degeneres have more Twitter followers than the entire populations of Ireland, Norway and Panama
Wow, so camwhores, twitter robots and secondary accounts for marketing transactions are people now? When will people learn that Free Account requiring only an email address != a person.

13. 80% of Twitter usage is on mobile devices…people update anywhere, anytime…imagine what that means for bad customer experiences?
Fun fact: First half of this statement is a statistic. Second half is speculation. LEARN TO TELL THE DIFFERENCE.

14. Generation Y and Z consider e-mail passé…In 2009 Boston College stopped distributing e-mail addresses to incoming freshmen
So one college, recongizing that many students already have an account and therefore don't need an additional one, represents and entire generate considering a tool is passe? That same tool that above counted into your social networking usage for the same generation? Really? And your source is "Metro Newspaper"

15. What happens in Vegas stays on YouTube, Flickr, Twitter, Facebook…
At least the source here stats: Opinion. Ugh.

16. The #2 largest search engine in the world is YouTube
Source is Which talks about Comscore. Interesting, but about a year old as far as stats go and is based on # of searches. Would be interesting to revisit post Bing-Yahoo deal.

17. Wikipedia has over 13 million articles…some studies show it’s more accurate than Encyclopedia Britannica…78% of these articles are non-English
Part 1: Stat. Part 2: NO SOURCE NOR FACT Part 3: Stat.
I guess 66% (see I can use stats too, cuz 3 data points is enough for a trend) stats is better than 0%. And this has to do with social networking again... how?

18. There are over 200,000,000 Blogs
*head aslope* Let's see, I've got my old G-blog, my MySpace blog (never used), my LiveJournal blog, a Blogger Account (never blogged there but still have a blog), my Facebook notes area (is that a blog?), my old blog from my Orko days, and I think I created a deadJournal and XJournal account. I also have a LiveJournal from my NanowriMo, and blog for my fanfic writings, a G-blog for DifferentVision and my GameSpot blog for my reviews. So I, by myself, have 12 distinct blogs. I worked on 2 more (worldcon in Boston Blog, blog for PrimoTech), so I'm up 16. Remember what we said previously about 1 free account is not strictly equal to a unique user? I'm sure there are more blogs I'm not even counting, and who knows what they are counting as blogs. Does my Twitter Feed Count as a blog? MEANINGLESS STATISTIC.

19. 54% = Number of bloggers who post content or tweet daily
... I'm just gonna leave this alone per the above, k? I don't even know how you begin to state that worldwide. Perhaps 54% of whomever Chris Aarons, Andru Edwards, Xavier Lanier interviewed for their "Turning Blogs and user-Generated Content Into Search Engine Results" article SES Magazine June 8 (WHAT YEAR WHO KNOWS) reported that. That can be generalized to worldwide stats, right?

20. Because of the speed in which social media enables communication, word of mouth now becomes world of mouth
Oh good another opinion. Aren't you clever. If you title your article Statistics, please leave your opinions at the door.

21. If you were paid a $1 for every time an article was posted on Wikipedia you would earn $156.23 per hour
And if I were paid $1 every time the article on Pope Benedict was updated for the first day I would have lots of money too. You mean that robots and people with easy access and the need to undo changes post a lot? Gosh. Wonder what people would have been paid per hour if you could have calculated the number of sentences written during the 1800s in Victorian England.

22. Facebook USERS translated the site from English to Spanish via a Wiki in less than 4 weeks and cost Facebook $0
The Source seems legit, though I doubt that $0 figure as Facebook setup the Wiki, paid the people to program and implement the voting system, paid to advertise the need, etc. Still, I have to admit coolness.

23. 25% of search results for the World’s Top 20 largest brands are links to user-generated content
Line out of the source I can't get from an earlier article. Again, how do you eve go about figuring out this stuff? And when does user-generated content begin and corporate induced repostings end?

24. 34% of bloggers post opinions about products & brands
I wonder how many conversationalists speak opinions about products and brands? WTF. There are entire blogs devoted to product/brand reviews and corporate sponsored ones. Is this really news that people talk about products and brands? Really? (Note the source on this is a video and the comments note that the methodology is not well defined).

25. People care more about how their social graph ranks products and services than how Google ranks them
WOW. Opinion phrased as something that could resemble a stat. I should do a stat on how many of these stats are really opinions.

26. 78% of consumers trust peer recommendations
Not news, not geared toward social media, and the exact question was 'have some degree of trust in X" Source (Author notes that it is 90% for a peer now). Further notes for this: Survey was online, which already is a self-selecting group, and many surveys note the skew for trust and technology up tick as a result).

27. Only 14% trust advertisements
Same source as above, and again these are not new stats for the last 15-20 years. OH WAIT NO IT IS A DIFFERENT SOURCE. Because the source above actually shows that 60% trust TV, Newspaper, or Sponsorships and over 50% trust Radio, Magazine, Billboard, etc. So you chose a different Source, that is a book geared to encourage people to market over social networks....

28. Only 18% of traditional TV campaigns generate a positive ROI
See above as the source is the book. Because books that are written with Bias and not peer reviewed are cool.

29. 90% of people that can TiVo ads do
Source is from TiVo which markets based on the ability to skip ads (which is what I think this statement means). People also fastforwarded commercials when they video taped things. And again I ask: What does this have to do with social networking? (I know the answer: it is that showing other things on the decline makes people assume your market is doing better, but again MISLEADING).

30. Hulu has grown from 63 million total streams in April 2008 to 373 million in April 2009
A ~500% increase in offerings year over year is good. One of the other major changes is that Hulu went from offering mostly long form content on a permanent basis to offering short clips, limited time content. Also, Hulu had been public for 1 month in April 2008 and was 1 month and a year old in April 2009. Note: Hulu is not Social Networking. Just in case any one forgot that's what we're taking about here.

31. 25% of Americans in the past month said they watched a short video…on their phone
Source is not able to be accessed, who know what month this was or when the study was conducted. Also, I really love seeing source data because it could reveal that they mean 25% of Americans with a mobile phone. Or 25% of Americas over 18. Or something.

32. According to Jeff Bezos 35% of book sales on Amazon are for the Kindle when available
So really what you mean there is that if a book is on the Kindle (of which we don't provide stats) 35% of the sales of those books are to the Kindle device. That is pretty cool, but look at the way the stat is phrased. "35% of book sales on Amazon are for the Kindle..." Not many will read those last two words or really understand the significance.

33. 24 of the 25 largest newspapers are experiencing record declines in circulation because we no longer search for the news, the news finds us.
Let's review this one more time: First part is Fact (though the source is not linked). Second part is speculation. Or Opinion. Be very careful with the use of the word 'because' in a stat discussion. Correlation != Causation. There's also this rather large economic downturn and the overall newspaper industry has moved to single paper towns which is close to a monopoly and has very little do with the internet or whatever you mean with 'the news finds us'.

34. In the near future we will no longer search for products and services they will find us via social media
An opinion that is a linked opinion. That's a source right? *STAB*

35. More than 1.5 million pieces of content (web links, news stories, blog posts, notes, photos, etc.) are shared on Facebook…daily.
Unverified #s from Facebook. Okay. And I'll buy it if you are including status updates, because holy shit people update all the time. And at least this is a Stat. About Social Networking.

36. Successful companies in social media act more like Dale Carnegie and less like David Ogilvy Listening first, selling second
No Source. How do you even have the balls... Not a Stat either.

37. Successful companies in social media act more like party planners, aggregators, and content providers than traditional advertiser
No Source. How do you even have the balls... Not a stat either.


As I said: The basic premise of the article is solid. The generation who is not 20-30 and uses these things as a second breathing, need to understand the importance of them. But I don't think using misleading, incorrect, or otherwise bullshit stats is a good idea. Nor do I think one should masquerade opinions as facts.

But I will note:

I got this article through work Email, from our person who does a lot on our corporate websites and our Twitter.
I twitter'd in anger about just wanting sources.
I IM'd a friend about several salient points.
I wrote this up on a blog (not facebook because I don't want to start a shitstorm with my coworker).
I used Facebook, Wikipedia, Firefox and other online sources when composing this blog.
I may even comment it to the author back, as they asked for feedback, and that is online interaction feeding back on itself.



fallen42 at 2009-08-15 16:51 (UTC) (Link)
I hate when people misuse numbers :(.
rbus at 2009-08-16 17:12 (UTC) (Link)
i shall never quote stats in an entry if there is a chance you you might read. unless they are true and i have a solid source. or at least, maybe 23% of the time i won't.

this is why *real* researchers go back to original source documents, and as many of them as possible.

as a researcher, believe nothing. especially that which bolsters your own point of view.

zhaneel69 at 2009-08-16 19:04 (UTC) (Link)
I mind less people quoting from a single source in a passing LJ entry. But this was geared as fact, selling a book, and designed to be used to make a point, and so it got my ire up.
rbus at 2009-08-16 23:02 (UTC) (Link)
a little p.t. barnum is good for the soul. but if my boss comes to me with that book i'm gonna print your rebuttal and refuse to take action.
Previous Entry  Next Entry