Meta-analysis

A meta-analysis is a study based on studies. It's often used in medicine or professional services when you are trying to establish patterns in existing data and a large-scale study is not feasible or not appropriate for some other reason. A meta-analysis can be sounder because the researcher can cast a broader net.

The main problem is, you have to have enough studies to analyze. They have to be close enough in terms of objectives and methodology that you can standardize them for analysis. And you have to know enough about each study to be able to include them with confidence. That means you have to know clearly whether a given study fits the parameters of your study so that you can include it.

You also have to be satisfied that all the studies you want to use are independent of each other and not just different version of the same study published by different primary authors under different titles. And remember that academia is just as at risk from informational bubbles as the housing market is to housing bubbles.

Another problem is that many of the scientific studies you might need to analyze are proprietary and cost money or may not be available to you as a poor slobs stuck in the private sector. You can sometimes get access to JStore, the place that grants access to scientific studies, through your local or university library. For example, I am a member of the Free Library of Philadelphia (not totally free) which entitles me to order articles for free.

If conducting meta-analyses is appropriate for what you are doing and you can do one, go ahead. You can get a lot of value out of them for relatively cost and you don't reinvent the wheel. Just make sure you avoid the pitfalls.

Check Wikipedia. There's an entry on it which is generally useful and check the links.

Entrepreneurship is the life's blood of all my professional activities. It makes them go. And go read my mystery for free at wwww.kearneymusicschoolmurders.blogspot.com or buy it from Amazon.com more cheaply than you can print it out.

Meta-analysis

A meta-analysis is a study of studies. Say you have 15 studies you've identified that measure the market share for widgets. Instead of collecting your own data, you do an analysis of the subject using study results as your data points. It's a way of creating primary data out of secondary data.

During the recent election, you might have seen the "poll of polls." That was the networks way of "averaging" the various polls done because the results were so inconsistent. A way of clearing out the clutter.

Wikipedia, the free, on-line encyclopedia says: "In statistics, a meta-analysis combines the results of several studies that address a set of related research hypotheses." Read the entire article and follow its links. I conducted a Google search on this and found lots of stuff, too, so read them.

And post a comment. I'd like to know what you think.

You often see meta-analysis in medical research. Medical studies are often conducted with a very small number of participants, so a good meta-analysis is actually stronger than any one of the studies. But I don't think until recently that the role of informational "bubbles" or "cascades" has been sufficiently recognized. That is, 100 researchers, all of whom read each other's work, can manufacture a false consensus. But that's for another blog posting.

Entrepreneurship informs everything I do. For my entrepreneurial real estate blog, go to www.yourstopforrealestate.com/blog and for entrepreneurial writing and publishing, to www.kearneymusicschoolmurders.blogspot.com.
In short, knowing how to do literature searches and meta-analyses is a basic entrepreneurial skill, so you had better at least learn about it.