What methodology can be applied when the calculated sample size is more than the population?

Home Forums Methodspace discussion What methodology can be applied when the calculated sample size is more than the population?

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #697

    Hello guys, I need some clarification and suggestions on the appropriate methodology for my PhD dissertation. I am working on a comparative analysis of webcast videos from three international television organizations over a period of 3 years (2012-2014). Based on my population size (740), I need to analyze a sample of 255 videos; 29 videos yearly for each TV organization. The sample size was determined using the formula by Saunders, Lewis, and Thornhill (2009)

    The challenge is that the population size of the videos related to my study for 2013 and 2012 for one of the select TV is only 13 (8 videos for 2013, and 5 videos for 2012) and  I need to include that particular TV in my study. This means that for those years I will be deficient of 45 videos. Kindly advise.

    #703
    Stephen Gorard
    Participant

    There seems to be a problem with your definition of population or your calculation. Perhaps both. I suspect that the source you cite is the problem.

    Do you have a list of all population cases? How many are there?

    Are you selecting cases for your sample randomly? I suspect not because you also say you need to include something or other.

    If you are not creating a random sample, then how did you calculate the sample size? 

    Why not try judgement? 

    See chapter 5 in 

    Gorard, S. (2013) Research Design: Robust approaches for the social sciences, London: SAGE, ISBN 978-1446249024, 218 pages

    #702

    Thanks Steve, below is the distribution of the study population:

     

    2012

    2013

    2014

    Total

    TV 1

    91

    93

    199

    383

    TV 2

    8

    5

    149

    135

    TV 3

    44

    127

    470

    222

     The population size was determined by retrieving all the available videos on the subject of my study posted by the select international televisions within the scope (2012 -2014). I intend to select the sample size using systematic sampling technique. I’m concerned with the fact that TV2 has far less content for 2012 and 2013 as compared with TV1 and TV3. I have searched the site severely and that all I found.

    #701
    Nino Rode
    Participant

    OK, something is wrong with this table. No total should be smaller than any of the summands…

    Anyway, I don’t see why you don’t you sampe proportionaly? This would, at least in my oppinion, be more representative of your structure of population.

    #700
    Clive Sims
    Participant

    I don’t think that this is possible Chioma. It is certainly not possible to have a ‘sample’ that is bigger than the whole population, because then the so-called  ‘sample’ is the population. Perhaps for that TV you need to use a whole-population statistic in which case the sample and the population are identical.

    Never come across this problem before.

    Clive Sims

    #699
    Seth Ansah
    Member

    Your sample size should be equal to, or less than your population. If have the means (money, time and effort) to capture all cases, you can consider using a census, where your sample is equal to your population.

    #698

    Check the total number of the videos per year. Are the values the sum of videos per year?If yes, then how did you come up with the values 383, 135 and 222?i guess thats where the problem is..

Viewing 7 posts - 1 through 7 (of 7 total)
  • You must be logged in to reply to this topic.