Tuesday, May 4, 2010

Reaching Curacao 2010 - Media Survey Methodology and FAQ

Methodology and frequently asked questions before the event.

  • March 2010 was the 2nd time we conducted a survey about media usage as part of a Caribbean wide effort. The first was time was February 2009.
  • The same survey has been done in at least 10 Caribbean countries at the same time.
  • It's a telephone survey, among 320 respondents, representative of Curacao. We have made sure that each age group, sex and educational status is represented in roughly the same proportion as in Curacao. In addition, we have tried to make each group large enough so that we can base some realistic conclusion on the findings. If you only have the responses of 10 males between the age of 18-24, you cannot base much of a conclusion on that.
  • Respondents were selected from our database (panel) and from random telephone listings. Why not just telephone listings? Because increasingly people do not want to answer questions from 'strangers'. Increasingly, around the world, market researchers are moving towards 'panels'.

What did we ask?
  • newspapers most read the past month
  • radio and tv stations listened to/watched each time of day
  • ownership of several items, including computers, internet, cellphones, BlackBerry/Iphone
  • use of internet
  • use of Facebook and frequency
  • most visited local website and frequency
  • roads most travelled (2009)
  • mode of transportation (2009)
  • geographic areas most shopped (2009)
  • demographics: age, sex, education, economic status (working, retired, student, etc.), profession, size of household
What can we report?
  • all of the issues above by age group, sex and education. We have developed a system to determine Social-Economic-Class (SEC) based on some of the demographics above. We report that, instead of education.
How exact is the data?
  • At a sample size of 320, the margin of error is around 5%. That means, if we find that 35% of all respondents reads newspaper Z, you can be 95% certain that the exact number is between 30% and 40%. Not exact enough? Consider what you are basing your media placement decisions on now...
Why didn't we use a larger sample size to provide more exact numbers?
  • Cost. After Reaching Curacao 2010 The Event and The Report, we will see what the appetite is for this type of studies. If it is big, we will do larger surveys. We certainly hope it will be!
  • ROI. When a research agency puts 1000 radio&tv monitoring boxes in households in Holland, advertisers stand to gain 100x more from the exact data these produce, than they would in Curacao. Why? Because the Dutch market is 100x larger and conducting a survey in Curacao is not 100x cheaper than in Holland. We can let ourselves be demotivated by the limitations of a small market or we can try to make the best out of it.
  • For more information on this subject: Survey sample sizes in small markets

No comments:

Post a Comment