AllExperts > Marketing Research 
Search      
Marketing Research
Volunteer
Answers to thousands of questions
 Home · More Marketing Research Questions · Answer Library  · Encyclopedia ·
More Marketing Research Answers
Question Library

Ask a question about Marketing Research
Volunteer
Experts of the Month
Expert Login

Awards

About Us
Tell friends
Link to Us
Disclaimer

 
 
 
 
About Alex J. Caffarini
Expertise
I can answer questions on any of the following: * Determining the best marketing research method for your needs; * Conducting surveys (including questionnaire design); * Measuring the effectiveness of marketing promotions; * Determining market size or market share; * Data analysis; * Statistical modeling; * Sales or business forecasting; and * Market segmentation.

Experience
I have 15 years of marketing research experience across several different industries, including banking, insurance, retail, and non-profit.

Organizations
American Marketing Association

Education/Credentials
M.B.A. in Marketing and Quantitative Methods, and B.B.A. in Economics, both from Loyola University Chicago.

 
   

You are here:  Experts > Business > Marketing > Marketing Research > reliability of data gathered

Marketing Research - reliability of data gathered


Expert: Alex J. Caffarini - 6/10/2009

Question
hi Alex hope you doing good, my question is how can we conduct post market research techniques to test the reliability of the study conducted using different market research techniques.
thanks

Answer
Sarz,

I see two approaches you can use.

The first approach is to take a sample of some of the research questions from the past marketing techniques, and then contact a random sample of the respondents.  Tell them you're just doing a quick follow-up to the survey they took.  See how their answers to the follow-up questions compare to their equivalent questions on the survey they took previously.  Test for statistical significance.  If the differences between the pre and post are significant, then those survey questions may not be reliable.  The drawback to this process is that it can be time-consuming.

Another approach would be a "split-half" reliability test.  In this approach, you separate your survey respondents into two groups and compare the responses of one group to that of the other.  While this approach is much more expedient than the approach above, it is also much more prone to error, as dividing the group into two does not have the ability to reveal differences within subgroups.  In this case, I would recommend stratifying the sample so that demographically, both groups look identical.  At that point, you would proceed with your split-half test.

Add to this Answer   Ask a Question


 
User Agreement | Privacy Policy | Kids' Privacy Policy | Help
Copyright  © 2008 About, Inc. AllExperts, AllExperts.com, and About.com are registered trademarks of About, Inc. All rights reserved.