Deceiving Statistics

May 28th, 2008

I feel cocky tonight, so I’ll present you with deceiving stats for this web site. I intentionally cropped the graphs to have 400px width so that they won’t mess up the layout of the web site. Giving, of course, the sample containing the highest peak.

If you see a graph, do not conclude immediately. You need more information in order to understand the situation — the behavior of the sample. Statistics is widely used now to give us false information. Every morning you will hear these information in the news. Researchers taking down stats and most of us make facts out of these samples. That’s why I hate the medical field so very much because of its dependence to stats.

Below is a Wordpress blogstat graph. Even though I’m showing a sample with the higest peak, blogstat shows that this is a typical activity for this site. That’s an average of 1000 visits. Blogstat does not compute according to a visitor’s uniqueness. So you may think that I’m bored every day and since I have no life, I visit the site 1000 or so times in order to see a graph looking like this.

You thought wrong! Below is a graph from Google Analytics. It’s really a boring graph because it’s nearly constant (almost a straight line). The graph below (the Y-axis has an interval of 1000 visitors) is showing an average of 1000 UNIQUE visitors daily. That proves the graph above to be legitimate.

Why then did I entitle this entry “Deceiving…”? Well, I intentionally omitted relevant information. The information I omitted do not affect the result of the graphs above, but they do affect the understanding of how the site is doing.

Related Posts

• Independent Thinker
• We The Living
• CloudBook
• Weekend Away
• twopointsix Upgrade

Entry Filed under: Site Updates