Oil Attention

A short update on the attention that oil prices are generating. See previous posts here and here.

Comments

[Matt Alcock]

This trend is interesting however I question the accuracy of the conclusions you can make from this data given the relatively small sample set. although there are a fair number of blogs analyzed (78 million) what proportion have any focus on energy and the environment? And what proportion have actually updated in the last 4 months, from the data on blogpulse I believe this is small in percentage terms (unfortunately this data may only represent a hand full of blogs, it may not but how can we tell?) Relative comparison of these items in this way on such a small sample set of blogs cannot give accurate conclusions can it?

Thoughts? Matt

[Matthew Hurst]

Matt,

The only conclusion you should make from this data is that attention is steadily increasing. Some thoughts:

'relatively small sample set' - blogpulse does a pretty comprehensive crawl of the active blogosphere;

'what proportion have actually updated' - this shows a plot of posts over time, so I guess you are asking "are these posts just one off posts?" I don't think that matters - in fact, if the oil situation is such that it causes non-bloggers to start a blog just to discuss it, surely that indicates even more attention!

'a handful of blogs' - the highest peak (June 9th) was caused by 2, 573 posts (http://blogpulse.com/search?query=%28oil+OR+gas+OR+petrol%29+%2B%28price+OR+payment+OR+afford%29&start_date=20080609&end_date=20080609); the final peak (July 2nd) by 2, 099 posts (http://blogpulse.com/search?query=%28oil+OR+gas+OR+petrol%29+%2B%28price+OR+payment+OR+afford%29&start_date=20080702&end_date=20080702). Is this a handful of blogs?

Of course, without further analysis, one can't draw any conclusions about the relationship between the two trends that I have shown.

BTW, I'd be interested in getting a better picture of what it would take to convince you that there is an upward trend here. I suspect that the most fundamental criteria is the coverage of the blogosphere - is that right?

[Matt Alcock]

Apologies Matt

The above proves the sample sets concerned are large, although blog pulse covers a large amount of the blog sphere I wrongly assumed many blogs would be updated only occasionally and very few would be around the topics above. The data as above prove these are large sample sets, in the region of 2,753 posts a day. Which is far higher than I had wrongly assumed.

Now knowing the rough number of posts included in the above, the trend really does prove across the blog shpere posts on oil and gas have increased. I guess this is a visualization issue rather than anything else by looking at percentages you can see relative change but it's hard to know what sort of numbers (sample set size) your are analyzing.

Regards
Matt