Greets to all,
I'm doing research in two phases- phase one is cross_sectional, while phase two is RCT. I did calculated the sample size for phase one and phase two. Phase two will be building based on phase one results. In phase one the data will be collected from 14 hospitals to determine their knowledge and attitude and that has low level in knowledge and attitude will be assigned randomly to RCT. So in phase one the sample size will be 339 participants. so I need to pick 339 participants From 14 hospitals. Which formula should I use to determine how many participants per hospital ? .
I mean can I say
339÷14 hospitals= 25 participants per hospital I will pick up to answer the questionnaire or there's special formula..... If Yes hit me up please
Best regards
Because to a statistician your 14 hospitals would be called 14 strata I am going to suggest Chapter 5 of this book: https://www.google.com/search?q=Cochran+Sampling+Techniques+3rd&rlz=1C1CHBF_enUS874US874&oq=Cochran+Sampling+Techniques+3rd&aqs=chrome..69i57.22745j0j1&sourceid=chrome&ie=UTF-8 This is a bit Mathematical but it is likely the best single volume available . Another choice at a lower level would be: https://www.google.com/search?rlz=1C1CHBF_enUS874US874&ei=lTotXubHMs_btQb6_arACg&q=elem a entary+survey+sampling+pdf&oq=Elementa&gs_l=psy-ab.1.2.0i67l3j0j0i67l5j0.260460.267446..276519...1.0..3.347.3664.35j2j1j1......0....1..gws-wiz.....6..0i362i308i154i357j0i273j0i131.uSPXFwD2E7s
, the chapter on stratified sampling. More information and a calculator can be found here: https://www.google.com/search?rlz=1C1CHBF_enUS874US874&ei=qzstXteHM9TEtQaSv6LQCA&q=sample+size+calculator+for+stratified+sampling&oq=sample+size+calculator+for+stratified+sampling&gs_l=psy-ab.3..0j0i22i30l5j0i333l3.206942.261343..265110...1.2..4.343.6918.63j10j1j1......0....1..gws-wiz.....6..0i71j0i67j0i362i308i154i357j0i273.N1dwi-V9370&ved=0ahUKEwiXiYue26DnAhVUYs0KHZKfCIoQ4dUDCAs&uact=5
Notice that this sampling approach provides better variance estimates under this sampling plan if you use the sample statistics methods given in the books. This is NOT what ordinary stat packages do. You need to select a special sampling package to get full benefit of doing this. The available packages in R are discussed here: https://www.google.com/search?rlz=1C1CHBF_enUS874US874&ei=tTwtXpfYL4W4tAa4yrSIAQ&q=R+package+for+stratified+sampling+computations&oq=R+package+for+stratified+sampling+computations&gs_l=psy-ab.3...404658.431619..434752...0.2..1.183.2998.32j3......0....1..gws-wiz.......0i71j0i7i30j0i8i7i30j0i333j0i67j0j0i8i7i10i30j0i30j33i10j0i22i30j33i22i29i30j33i160.5MwQHxbtAjY&ved=0ahUKEwjXhvOc3KDnAhUFHM0KHTglDREQ4dUDCAs&uact=5 If you prefer another Stat package say SAS simply Replace R by SAS in the google command. Not all packages have this feature and it usually is at additional cost. Not so for R, of course. Best wishes, David Booth
Dear Awwad,
Please see the following link and I hope it will help you out:
Article Optimal Allocation of Sample Sizes to Multicenter Clinical Trials
Best of Luck
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size