“Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital.” ~Aaron Levenstein”
I had a post in mind today, one about metrics and statistics, some of which you may have heard if you were at my presentation . Before starting I Googled “quotes about statistic”, as I thought I wanted to use a starting quote that was better than the obligatory “98% of all statistics are made up”. I have discovered that quotes about statistics are bit like statistics themselves, many, and varied, they can be also both humorous and pointless. So I shall drop several more through the post as I had trouble deciding which I liked most.
“Torture numbers, and they’ll confess to anything.” ~Gregg Easterbrook
So why statistics? Well there are many different areas with which statistics are important in libraries, but for me the main statistical set that has me beating my head against a brick wall is blog statistics. It is true that with any activity within the workplace, their needs to be a justification for that activity. Often the easiest was to prove/ or disprove that justification is with statistics. For the blog, Tararua District Library Blog, statistics are any essential part of any justification.
“He uses statistics as a drunken man uses lampposts – for support rather than for illumination.” ~Andrew Lang
The trouble with the statistics are they are often inconsistent across harvesters and often they don’t tell me the story I want to know. I can start with the statistics generated by wordpress. They record how many times a particular page has been viewed in a browser. They don’t tell me a location of the reader, nor if the reader is a unique visitor who viewed multiple pages. So for example last month [October] I am told the blog had 1,420 page views, at an average 46 page views per day. This is well and good, but not terribly satisfying for those with an interest in metrics.
“Facts are stubborn things, but statistics are more pliable.” ~Author Unknown
I needed to see some more statistics, so I created a free account with sitemeter, which tells me both unique visitors and page views. Except this service tells me for the month of October we had 1,129 visits, viewing 1,592 pages. Yes that’s 172 page views more than the blog tells me. That’s more than a discrepancy in date/time [I.E. one site recording at GMT, while the other NZ local time]. One thing I would be like to be able to report is the numbers of locals reading. Sitemeter will tell me an approximate locale of the viewer, but that is based of the ip address, and they are usually registered in bulk to one of the main centres.
“There are two kinds of statistics, the kind you look up and the kind you make up.” ~Rex Stout, Death of a Doxy
To add to my confusion I have recently started channelling my links through bit.ly. Now this has told me some interesting statistics which is meant to report how many times the link has been clicked. The obvious conclusion would be that for every click you would receive a corresponding page view. I have discovered this is in fact not true. I can create a link to my latest posting, send it to twitter via bit.ly. I have observed that in bit.ly I can have 12 or 13 clicks yet not one page view.
“Satan delights equally in statistics and in quoting scripture….” ~H.G. Wells, The Undying Fire
To add another element in the quest for an accurate picture of readership, there are also blog harvesters like alphainventions.com or http://stumbleupon.com, which will allow people to read your blog without actually viewing it. Not to mention the question of whether you can get an accurate readout of how many people have subscribed via a RSS feed.
It means that any validation of the blog as a worthwhile activity, justified as active part of the library workload is problematic. I feel sometimes as though I am out there, dancing in that bikini on the street front, not knowing how many are checking me out, and living in fear that the powers that be might just notice and decide I need to be engaged in a more appropriate activity.
One reason for the bit.ly discrepancy is that sites like FriendFeed often expand the bit.ly link in their display (so people can see the source before clicking on it) – this registers as a bit.ly stat even if no-one clicks on the link to give you a blog stat.
I imagine the discrepancy between WordPress and Sitemeter is that WordPress isn’t counting some kinds of hits – might be spiders, might be spambots, might be hits from your own IP address.
That explains a lot re: bit.ly, I wondered if that might be the case..
WordPress automatically ignores your own hits, while I have Sitemeter set to ignore them as well.
But yes there are a lot of different variables.