How does one calculate a trends (up 7%, down 7%, etc.) on a series on numbers?
I'm new to statistics and trying to learn how to discover a trend in my website visitor totals.
So for example I have 30 values, one for each day of the month. I know how to calculate the standard deviation, but how to I measure whether or not my visitors are increasing since the data is somewhat random and will vary slightly on a daily basis. This is keeping in mind i don't have a previous month to compare against and simply want to measure growth on this 30 day period. I don't want to simply compare the first day to the last as they may be off days.
This Question has been solved and asker verified All Experts Exchange premium technology solutions are available to subscription members.
Experts Exchange has been collecting answers to technology questions since 1996…3 million and counting! If you have a question, chances are we already have your answer.
If you can't find the exact answer you're looking for, ask our exclusive community of 50,000 experts. You’ll get a personalized answer from a trusted professional.
Thousands of free tech tips, tricks, how-to’s and tutorials are available in our peer reviewed articles section. See for yourself how smart our experts are, no login required.
Save solutions to your questions, answers you’ve discovered through searching plus helpful articles in your personal knowledgebase for easy future access.
Access the answers to your technology questions today.
30-day free trial. Register in 60 seconds.
Members of the expert community talk about why the experience at Experts Exchange is different than what you will find anywhere else.
Try it out and discover for yourself.
30-day free trial. Register in 60 seconds.
Join the community of experts here and help other tech pros by answering question in your area of expertise. You can earn FREE access to all Experts Exchange's premium features and resources.
Well, you can calculate % change over whatever interval you want... Computing an average on the last 3 days and the first 3 days will compare just that - how does traffic at the end of the month compare with that at the beginning of the month. You don't really have enough data to do much else.
As ozo suggested, you could also use regression - the slope from that statistical procedure would tell you the average number of users that you increased each day over a certain time period, accounting for daily fluctuations, at least to some extent... if that's what you're interested in?
Yes I think the linear regression is just what I need. I found this formula online:
http://people.hofstra.edu/
So would I just do the plots for each X value to get my trend line and then compare the percentage of difference between the first and last plot?
Not precisely, but that's close. The line of best fit is produced entirely mathematically - you need to compute the slope and intercept to best fit the data, so there's not really any plotting involved. You'd first convert your days to simple numbers (i.e. 0 to 31) and use that as the predictor (x). You'd end up with a formula like: y' = day*slope + intercept
Then, you'd use that formula for values at both the beginning of the month:
beginning_trend_value = 0*slope + intercept
end_trend_value = 31*slope + intercept
Then you could use this formula:
%gain = end_trend_value / beginning_trend_value * 100
You could substitute any value for the 0 and 31 if you wanted to know the trend across any specific days.
If you want to know how to compute the regression terms, I can step you through it (I teach statistics at a University in the Midwest), but this page (http://phoenix.phys.clems
Business Accounts
Answer for Membership
by: richdiesalPosted on 2009-01-26 at 22:34:16ID: 23474121
Unfortunately, there's no simple answer. Rolling averages are probably the most common method - compute weekly averages or even 3-4 day averages and then plot those averages on a graph with day on the x-axis. If you plot it with one bin size (say, 3 day), and you see a spike or drop, then compute it with another bin size (or two) and see if that spike/drop is still there.
Does that help?