Enter your name and e-mail address to quickly register and be notified of new articles.
Keep Me Updated!

one moment please...


Blog Home Tags: Print

A New Study Suggests...

image loading... by Bo Bennett, PhD, Social Scientist, Business Consutlant
posted Friday May 24, 2013 12:00 AM

image loading...

Bo Bennett, PhD

Social Scientist, Business Consutlant

About Bo Bennett, PhD

You can read my full bio at http://www.BoBennett.com.

"A new study suggests..." just about anything and everything. There are many reasons for this ranging from flawed methodology to outright fraud. One concept that probably confuses the public most is generalizability, or drawing conclusions from one or more studies to an entire population.

Let's say we are comparing men and women and their coin flipping abilities. In our tests, the men flip heads 51.8% of the time. The women flip heads only 49.8% of the time. The results are published, and the headlines in the media read "A New Study Suggests Men Are More Likely Than Women To Flip Heads When Tossing Coins." In this case, the results are almost certainly due to chance (statistically insignificant results) and would not hold true for the population at large.

My advice:

1) If you see a "shocking" headline about the results of the study, READ THE ARTICLE! Any media outlet that as at least some set of standards will make it clear in the text of the article (usually at the very end) how insignificant the findings actually are—assuming they have read the study!

2) Google the study and look at how other media outlets reported the results (or if they even did). If they didn't, it is very likely that there is nothing "shocking" going on, just inaccurate reporting. If there are conflicting spins on the data, you know someone is misinterpreting or spinning the data. Is it possible that your news source (mypoliticalparty-quote-news-endquote.com) is?

3) Find the original paper yourself using Google Scholar (scholar.google.com). You can almost always access the abstract for free and quickly get a good idea of the accuracy of the claims being made.

In short, it is not a good idea to make life-changing decisions based on headlines. Do a little research first. It's worth it!


---------------------------------------------------------------------
Bo Bennett, PhD
Founder and Developer of http://www.YourWebEmpire.com — A modular, web-based platform designed to quickly and easily create multiple attractive and highly-functional web properties.

The Dr. Bo Show > http://www.TheDrBoShow.com
About Me > http://www.bobennett.com
Private, Anonymous Comment On This Post (no login required)Your comment below will be anonymously sent to the post owner, it will not be posted, and you will not get a response. To make a public comment, post below (login required).

Send Comment sending comment...

Registered User Comments




 Copyright 2017, Archieboy Holdings, LLC. 

Component Viewer

A component is the HTML code for a section of a webpage that can be combined with other components to make a complete webpage. Click the component to insert the component code at the bottom of your current page, then customize it.