Thanks! These are wonderful for your very detailed explanations. You take terms that people often gloss over and really describe them in depth, such as reliability and validity, etc. I am loving all these videos and working my way through them. Thank you so much for this!!
Thanks U for ur sharing...ur videos makes me clear about my lessons and If ur have more videos , plz upload because I am studying research methodology.Another one is ur videos save my time.
Andreas, I'm sorry I can't. I heard about it many decades ago; I think from reading Claus Moser and Graham Kalton, Survey Methods in Social Investigation (Ashgate Dartmouth 1985), but I don't have my copy to hand so I can't check. It was certainly a very old study probably from the 1950s or 60s.
Hi Graham, How many people should be selected in order to get accurate result for the subject that you are studying. I am thinking of launching a product. I think I might sound crazy but feel I should get 10k people to participate in order to get accurate reflection.
It is rarely necessary to take that large a sample (unless you want to analyze some small subgroups). If you have some information about the variables you are measuring (effect size, variance) then you can calculate the necessary sample size needed for a given power (see: Ellis, Paul D (2010) The Essential Guide to Effect Sizes. Cambridge). But typically we do not have such information (or an estimate of it) before the survey. Also, how accurate is accurate? Social scientists usually use an alpha of 0.05. But if you intend to invest lots of money, you might want to use an alpha of 0.01. This effectively widens the error bounds in your estimate of the population characteristics.
Husam BinSasi This simple formula is somewhat contentious but it can be used in very simple cases. It assumes we are using a confidence coefficient of 95% and we are trying to estimate the proportion of a population in one variable (e.g. the percentage of people supporting leaving the EU) and that this is close to 50%. But in most cases we are dealing with many variables and we want to look at sub-populations too, so the figure it produces may be misleading. Most calculations of sample size, such as Cochran's require an estimate of the variance of the variable we are investigating in the population concerned. Slovin's formula: n = N / (1 + Ne2) where n = Sample size N = Total population and e = margin of error.
You can make money from this website just by listening to music and reviewing the song. It's not a lot but it is legit. You make about $20-70 Per day epending how long and unique your review is. Start here! *ONLINEGETCASH .INFO* *ONLINEGETCASH .INFO*
Absolutely EXCELLENT chief Gibbs, many people like me had benefited on a daily basis, please keep doing it and thank you very much
Excellent and very informative.
Thanks! These are wonderful for your very detailed explanations. You take terms that people often gloss over and really describe them in depth, such as reliability and validity, etc. I am loving all these videos and working my way through them. Thank you so much for this!!
I was so lost as a grad student. Watching this helped me. I have a few questions and was hoping you can help....
Great video for postgraduate statisticians. Thank you very much for uploading it.
Thanks U for ur sharing...ur videos makes me clear about my lessons and If ur have more videos , plz upload because I am studying research methodology.Another one is ur videos save my time.
Andreas,
I'm sorry I can't. I heard about it many decades ago; I think from reading Claus Moser and Graham Kalton, Survey Methods in Social Investigation (Ashgate Dartmouth 1985), but I don't have my copy to hand so I can't check. It was certainly a very old study probably from the 1950s or 60s.
Hi Graham, How many people should be selected in order to get accurate result for the subject that you are studying. I am thinking of launching a product. I think I might sound crazy but feel I should get 10k people to participate in order to get accurate reflection.
It is rarely necessary to take that large a sample (unless you want to analyze some small subgroups). If you have some information about the variables you are measuring (effect size, variance) then you can calculate the necessary sample size needed for a given power (see: Ellis, Paul D (2010) The Essential Guide to Effect Sizes. Cambridge). But typically we do not have such information (or an estimate of it) before the survey. Also, how accurate is accurate? Social scientists usually use an alpha of 0.05. But if you intend to invest lots of money, you might want to use an alpha of 0.01. This effectively widens the error bounds in your estimate of the population characteristics.
Excellent.Thank you 💐
What about Slovin's Formula?
Doesn't it look like a very useful way to figure out the size of population we want to study?
Husam BinSasi This simple formula is somewhat contentious but it can be used in very simple cases. It assumes we are using a confidence coefficient of 95% and we are trying to estimate the proportion of a population in one variable (e.g. the percentage of people supporting leaving the EU) and that this is close to 50%. But in most cases we are dealing with many variables and we want to look at sub-populations too, so the figure it produces may be misleading. Most calculations of sample size, such as Cochran's require an estimate of the variance of the variable we are investigating in the population concerned.
Slovin's formula: n = N / (1 + Ne2) where n = Sample size N = Total population and e = margin of error.
Thanks
thank you
continue excellence
You can make money from this website just by listening to music and reviewing the song. It's not a lot but it is legit. You make about $20-70 Per day epending how long and unique your review is. Start here! *ONLINEGETCASH .INFO*
*ONLINEGETCASH .INFO*
Can you give us any more info about that wedding ring survey?
thank you very much
Thank u!
We appreciate your feedback. Participate here www.surveymonkey.com/r/82CZZ88
16 20