November 24th, 2008 · 3 Comments · 43 views
When data are missing, the appropriate missing data analysis procedures do not generate something out of nothing but do make the most out of the data available
One of the most common forms of analysis with missing data involves simply substituting the mean for the variable whenever a value is missing. Unfortunatley, mean substitution can produce very wrong estimates of variances and covariances. In general, substituting the mean for the missing value has the effect of underestimating the magnitude of both variances and covariances
“1. Whenever possible, use the EM algorithm (or other maximum likelihood procedure, including the multiple-group structural equation-modeling procedure or, where appropriate, multiple imputation) for analyses involving missing data.2. If other analyses must be used, keep in mind that they produce biased results and should not be relied upon for final analyses. Recommmending:
a. Never use mean substitution, even for preliminary analyses.
b. With minimal missing data, analysis of complete cases may be a reasonable solution.
c. If data are missing completely at random, pairwise deletion or complete cases analysis may be a reasonable solution.
d. If data are not missing completely at random and the cause of missingness has been measured, complete cases may produce unbiased e2stimates, although it is a generally less powerful approach than the EM algorithm or multiple-group procedure.”
John W. Graham, Scott M. Hofer, and Andrea M. Piccinin (1994). Analysis With Missing Data in Drug Prevention Research L. M. Collins & L. A. Seitz (eds.), Advances in Data Analysis for Prevention Intervention Research (13-63). NIDA Research Monograph 142 Bethesda, MA: U.S. Department of Health and Human Service
Accordingly, mean substitution:
1. artificially decreases the variation of scores, in turn, this decrease in individual variation for each of the variables is proportional to the number of missing data - in turn, the more missing data, the more “perfectly average scores” will be artificially added to the data set; and
2. substitutes missing data with artificially created “average” data points - this can result in considerably changing the values of correlations.
We have tried to minimize this issue with calculating the mean value for those variables with the same Google Page Rank only (e.g., take all means from variable x for those blogs with Google PageRank 4 only - calculate the average score - use it). In turn, this reduces the impact outliers - high and low - may have on our results. For more see also:
what we do with missing values
We are currently in the process of finding ways for doing multiple - imputations. Want to support our efforts, get in touch and/or leave a comment below.We are wondering how we can integrate this program:
Schafer, J. Software for Multiple Imputation
with our work. Suggestions are welcome.
We need to find a way to run it without using any of the statistical packages to do the job for us - any advice - please leave a comment we need your expertise and appreciate any help we can get.
other resources
Allen, E. I., & Sharpe, N. R. (2005) Demonstration of Ranking Issues for Students: A Case Study. Journal of Statistics Education Volume 13, Number 3
Karen Grace-Martin - writes The Analsis Factor blog - you should subscribe - it is refreshing and very helpful indeed
Posted in ranking | 3 Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking
October 6th, 2008 · 1 Comment · 140 views
Why would you want to be an alpha tester - if your answer is yes go here: click here to register yourself now
So you want to be convinced? We could try by writing some fancy text. Instead how about asking you four questions that you should answer for yourself right here:
1) are you a blogger?
If your answer is yes, read on.
2) do you own or are responsible for a blog?
If the answer is yes, read on.
3) do you want to compare your performance against how well other bloggers are doing?
Probably your answer is yes. You want to see how you benchmark and rank against others.
4) want to see how your performance changes over certain periods?
Of course, you want to know how things change over time (e.g., first Quarter against fourth Quarter 2008).
Most importantly, how the First Quarter in 2009 is comparing to your First Quarter in 2008 you want to know.
How about a tool that visualizes these changes and compares things to your closest competitor? You can get it.
so what are you waiting for?
Answered yes to the questions listed above, go register and take us for a test-drive.
Just remember, we are still developing this tool and things are improving as you read this post - which means new services, tools and tricks added every other week it seems.
Nevertheless, start tracking your blog’s performance using this information to become more effective with your blogging efforts.
click here to register yourself now
By registering you can benchmark and compare your blog right away. Of course, alpha testing is free, what else.
Posted in ranking | 1 Comment »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking
September 8th, 2008 · No Comments · 219 views
So you have claimed your personal blog and now you wonder whom to compare yourself to (where am I better?), to see how well you are doing.
Here are some ideas. You may follow these suggestions. As well, if you have some ideas that helped you but you cannot find them in the list below, please share them with us and leave a comment.
1) apples against apples
If you write a travel blog it makes sense to compare yourself to another travel blog.
2) gorilla against gorilla
So if your blog has a huge following, why not compare yourself to another gorilla?
Even if your blog is small, you may compare yourself to one gorilla type of blog that you like very much. This will get you a feel about how well you measure up compared to the guys and gals in the big leagues. So go ahead and trace that gorilla-type blog
3) compare yourself against the most appropriate set of competitors
Some people blog under their own domain which I recommend as the best way to build one’s own brand.
Others may have their blog at a host, whereby their URL may look like similar to this:
http://blog.ebookers.ch/author/wmueller/
It might be interesting to compare oneself to other blogs under this domain, such as:
http://blog.ebookers.ch/author/aroberts/
It might even make sense to compare yourself to the overall blog of this blogger host, such as:
http://blog.ebookers.ch/
Nonetheless, it is helpful to compare yourself also to blogs addressing the same topic that you do that are, however, hosted under their own domain.
4) rule of 10
It is probably a good idea to compare one’s blog with a set of about ten blogs.
This set of 10 blogs that you track and compare with might include a gorilla or two, a colleague or friend and so forth. As well, one or two blogs that you admire could be included in this set of 10 blogs. In fact, these two blogs may cover a different subject than what you write about. Nevertheless, because you like them - maybe even subscribe to their posts - it could be a good idea to track and compare your blog with these. Learn how they measure up compare to your own blog.
At the beginning, comparing oneself to a set of 10 blogs might suffice. Thereafter, you may want to track more blogs enabling you to get a more accurate picture of where you stand in the blogosphere.
4) it took more than 1 day to build Rome or Paris
It takes time. Don’t try to come to any conclusions within 30 days. In fact, it will take at least four months before you can see some trends that might indicate that your blog is improving in the rankings.
Look at the graphics produced at My.ComMetrics.com, ask for a report or just compare yourself and watch the trends.
Rome was not built over night, neither will your blog be a two week project. So take the time, post regularly, provide people with beef and watch how your numbers improve using My.ComMetrics.com to benchmark yourself - easy does it.
By Invitation
At this point you can test our Beta version by invitation only. However, if you send us a note using this mechanism, we will send you an invite:
get a guest account and track your blogs - send a note - get invited
Else, get an invite from one of our current subscribers right now. Another fast was to get access to our tools now instead of tomorrow.
Posted in ranking | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking
August 22nd, 2008 · No Comments · 361 views
When you look at successful web applications, they are all simple to use. Moreover, users benefit from using the tool nearly instantly, such as:
Google, type in your keywords, search and find what you are after;
Plaxo, get a username and password ==> start adding your addresses and synchronize with your PC, and NOW
Of course, Plaxo takes a bit more time to make it work properly on your PC. Nevertheless, its address synchronization feature between your PC and Plaxo works right away after having installed a small plug-in. Other features make it easier to better manage your social network.
When you have to read too much to figure out what ingredients are needed to cook a dish, forget it.
All it takes is a good bread recipe and some flower, yeast, water, salt. Follow the recipe and you are pretty much ready to make the dough, let it rise, put it in the oven …… The result is a delicious bread for you to share with your family and friends.
Good ideas help making you do whatever you need to do. Sure, some web-based applications are a bit more difficult than Google or Plaxo as described above. Nevertheless, they are easy to use, such as Onvista.de allowing you to:
a) track certain stocks, after you;
b) register and get password; thereafter
c) enter the stocks you want to track; AND
d) see the results.
It will take about 30 minutes to familiarize yourself so much that you can really take advantage of the service.
Same way things work with My.ComMetrics.com
All it takes is for you to sign up, register your blog, enter those blog URLs you want to compare yourself against and you are set.
fill in contact form, call or e-mail us to get invited as an alpha user RIGHT NOW.
Posted in ranking, social media measurement | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking · social media measurement
July 12th, 2008 · No Comments · 230 views
In an ever faster changing global marketplace, the Internet and social media are being used by stakeholder groups to connect to each other. Often this happens without the direct involvement by the corporation owning the brand.
Today, a more effective spending of hard euro marketing costs is needed. Furthermore, it is critical that social media activities should help increase the firm’s lead generation and customer retention.
Such efforts can help in reaching potential clients and embracing existing customers by creating communities. While the company cannot respond to every comment people might leave on a blog or micro-blog to the CEO, social media helps the listening process regarding investors’ and customers’ needs and concerns.
An ever greater number of people is getting excited about Web 2.0 technology. One of the greatest challenges is explaining why this all matters, and how this all works, to people who have no idea what the technology can or cannot do.
Download the version as a pdf file here:
Financial Times - Is blogging good value for the C-suite? 2-pages
Our tools empower you to benchmark social media efforts and activities, while focusing attention on customers, trust and confidence, community, and social connections, interested?
Or read the article online here: Read more on the Financial Times website
What do you think?
Posted in ranking, social media measurement | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking · social media measurement
June 23rd, 2008 · No Comments · 827 views
Measuring impact with only a single coarse tool when it comes to social media is dangerous
We look at various statistical indicators to see whom the crowd is following on the Internet and what this means for the brand.
There is no single metric for ranking the popularity and/or influence of your Chief Executive Officer’s (CEO) blog, your company’s page on Facebook, monitoring Twitter conversations and so on. There are two principle ways of ranking social media:
- traffic (which is not a term with automatic meaning) and
- through the number of links to a page - thes have generally been used as sign of influence.
Nevertheless, blog content depends on the quality of the blog or the social media channel in which the original story or post appears. These, in turn, depend on the integrity of the writer who checks the facts and makes sure that the material is correct.
Nevertheless, in the social media domain,
have become ever more important when assessing the influence of a blog or a social media campaign.
Essentially, similar to citation analysis for research, most scoring systems assessing the popularity or backlinks for a blog are focusing on the number of times a blog post is linked to or referred to by another blogger.
However, a report by the International Council for Industrial and Applied Mathematics, the International Mathematical Union and the Institute of Mathematical Statistics states in a recent press release to a report warning over science citations:
While having a single number to judge quality is indeed simple, it can lead to a shallow understanding of something as complicated as research. Numbers are not inherently superior to sound judgments.
Get the full report here:
Measuring Twitter, MySpace, Bebo, Facebook & blogs with a single metric is dangerous (416.9 KB)
Posted in Uncategorized | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: Uncategorized
May 30th, 2008 · No Comments · 278 views
We are developing our tool and testing first results
Below we give you a glance at how it is looking
We have been focusing on some of the popularity measures to address benchmarking in the context for blogs, webpages and mini-blogging (e.g., Twitter).
Below is a first screenshot from the dashboard that we offer our Alpha users so far.
If you cannot see the above screenshot try - crowdscoring your social media efforts - ComMetrics tool
Naturally, data presented in a table such as the one above have to be added and normalized to arrive at the compositive index. As well, there are a set of compositive indices that you can use to benchmark and rank yourself.
The dashboard enables the user to decide if things are listed in ascending or descending order, if she would like to see raw data or normalized data instead or both.
Naturally, the preference given to which indicators one wants to look at is another option.
Bottom Line
We continue to work on this to make it work even smoother than it does now. As well, we are putting in the visual options that allow users to get a trend analysis. In the latter, comparing one(s) blog(s) or webpage(s) with a set of possible competitors or best practic examples is a feature you can look forward to.
So stay tuned.
Posted in ranking, social media measurement | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking · social media measurement
April 25th, 2008 · No Comments · 557 views
A few of our clients and readers have asked to hear what we have been up to and when the alpha version will finally go Beta.
Well we are in the process of inviting a few of you to pre-test the Beta version. If you have subscribed to the ComMetrics blog - e-mail digest you might just get invited directly enabling you to join the small group of knowledgeable users like yourself that have been testing our tool for the last 60 days or so.
At this stage we have gone into private beta. Nevertheless, we should be ready to invite our ComMetrics blog subscribers in a few weeks to start with their test-drives.
In the meantime, we keep you posted here as well as on our ComMetrics blog - stay tuned.
For now, no panic we have started to do a more formal write-up of what we have in store for you that you can look at using these hyperlinks (see also right side - pages - of website - more details about what we are into - real fun and hard work)
To learn more contact CyTRAP Labs
Posted in ranking, social media measurement | No Comments »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking · social media measurement
April 11th, 2008 · 1 Comment · 325 views
not every CEO wants to blog - but some do better than others
My view on most blog rankings is quite simple: I think they’re trying to compare and rank things you cannot really compare and rank.
Many depend on the number of links in Technorati or the number of visitors. There are numerous variations of these and we have tried to outline this here:
Popularity - who is the favorite to win
Unfortunately, this starts on the premise that blogs target the same audience and provide the same type of information. We have pointed out elsewhere that the key is to identify your audience and the niche for your blog or the content you wish to put up on a website.
We can compare apples with apples but should refrain from comparing oranges and apples.
So if we cannot just compare blogs or website since they are not the same, what is the second best thing to do?
Well one can go ahead and compare sports cars with sports cars, pickup trucks with pickup trucks and trying to rank them in order to find out which one might be the best. That still does not address if we are trying to find the most economical, nice looking or greatest workhorse amongst the trucks chosen.
This is one reason why we are trying to rank the CEO blogs according to industry and so on…..
top CEO blogs
Posted in ranking | 1 Comment »
Related Posts:
Email This Post
|
Print This Post
Tags: ranking
