پديد آورندگان :
روﺷﻨﯽ, دائم داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﮑﯽ ﮐﺮدﺳﺘﺎن - ﻣﺮﮐﺰ ﺗﺤﻘﯿﻘﺎت ﻋﻮاﻣﻞ اﺟﺘﻤﺎﻋﯽ ﻣﺆﺛﺮ ﺑﺮ ﺳﻼﻣﺖ - ﭘﮋوﻫﺸﮑﺪه ﺗﻮﺳﻌﻪ ﺳﻼﻣﺖ , نوري, بيژن داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﮑﯽ ﮐﺮدﺳﺘﺎن - ﻣﺮﮐﺰ ﺗﺤﻘﯿﻘﺎت ﻋﻮاﻣﻞ اﺟﺘﻤﺎﻋﯽ ﻣﺆﺛﺮ ﺑﺮ ﺳﻼﻣﺖ - ﭘﮋوﻫﺸﮑﺪه ﺗﻮﺳﻌﻪ ﺳﻼﻣﺖ , ﻣﺮادي, ﻣﺴﻌﻮد داﻧﺸﮕﺎه ﻋﻠﻮم ﭘﺰﺷﮑﯽ ﮐﺮدﺳﺘﺎن - ﻣﻌﺎوﻧﺖ ﺗﺤﻘﯿﻘﺎت و ﻓﻨﺎوري
چكيده فارسي :
ﯾﮑﯽ از ﻣﺴﺎﺋﻠﯽﮐﻪ اﮐﺜﺮﻣﺤﻘﻘﯿﻦ در ﺷﺮوع ﭘﮋوﻫﺶ ﺧﻮدﺑﺎ آن ﺳﺮوﮐﺎر دارﻧﺪ، ﻣﻮﺿﻮع ﺑﺮآوردﺣﺠﻢ ﻧﻤﻮﻧﻪ و اﻧﺘﺨﺎب ﯾﮏ ﻧﻤﻮﻧﻪ ﻣﻨﺎﺳﺐ از ﺟﺎﻣﻌﻪ ﻣﻮرد ﻣﻄﺎﻟﻌﻪ اﺳﺖ. ﺑﺮآورد ﭘﺎﺋﯿﻦ ﺣﺠﻢ ﻧﻤﻮﻧﻪ ﺳﺒﺐ آﺳﯿﺐ ﺑﻪ اﻋﺘﺒﺎر ﭘﮋوﻫﺶ و ﻫـﺪر رﻓـﺖ ﻫﺰﯾﻨـﻪ و زﻣـﺎن ﻣﯽ ﺷﻮد در ﻣﻘﺎﺑﻞ ﺑﯿﺶ ﺑﺮآورد آن ﺑﺎﻋﺚ اﻓﺰاﯾﺶ ﺣﺠﻢ ﮐﺎري، ﻫﺰﯾﻨﻪ و ﻋﺪم ﺑﺎزﺧﻮرد ﻣﻮرد اﻧﺘﻈﺎر ﺧﻮاﻫﺪ ﺷﺪ.ﻫـﺪف از اﯾـﻦ ﭘـﮋوﻫﺶ ﺗﻌﯿﯿﻦ ﺣﺠﻢ ﻧﻤﻮﻧﻪ در ﻣﻄﺎﻟﻌﺎت ﻣﺨﺘﻠﻒ ﻋﻠﻮم ﭘﺰﺷﮑﯽ ﻣﯽ ﺑﺎﺷﺪ.
ﻣﻮاد و روش ﻫﺎ: در اﯾﻦ ﻣﻄﺎﻟﻌﻪ ، ﭘﺲ از ﺗﻌﯿﯿﻦ اﻧﻮاع ﻣﻄﺎﻟﻌﺎت، در ﻣﻨﺎﺑﻊ ﻣﺨﺘﻠﻒ داﺧﻠﯽ و ﺧﺎرﺟﯽ )ﮐﺘﺎﺑﻬﺎ، ﻣﺠﻠﻪ ﻫـﺎ، ﻣﻘـﺎﻻت و ﺟﺴﺘﺠﻮ اﻧﺠﺎم ﺷﺪ و روﺷﻬﺎي ﻣﺨﺘﻠﻒ ﺗﻌﯿﯿﻦ ﺣﺠﻢ ﻧﻤﻮﻧﻪ ﺑﺮ اﺳﺎس ﻧﻮع ﻣﺘﻐﯿـﺮ، ﻫﻤـﺮاه ﺑـﺎ ﻓﺮﻣﻮﻟﻬـﺎي ﻣﺮﺑﻮﻃـﻪ اﺳـﺘﺨﺮاج و دﺳـﺘﻪ ﺑﻨـﺪي ﺷﺪﻧﺪ. از ﺑﯿﻦ روﺷﻬﺎي ﻣﺨﺘﻠﻒ ﺑﺎ ﺗﻮﺟﻪ ﺑﻪ ﻧﻮع ﻣﻄﺎﻟﻌﻪ و ﺳﺎﯾﺮ ﻋﻮاﻣﻞ ﻣﻬﻢ و ﺗﺎﺛﯿﺮﮔﺬار، ﺑﻬﺘﺮﯾﻦ ﻓﺮﻣﻮل و روش اﺳﺘﺨﺮاج ﺷﺪه اﺳﺖ. ﯾﺎﻓﺘﻪ ﻫﺎ: ﭘﮋوﻫﺶ ﻫﺎي ﻣﺨﺘﻠﻒ ﻧﺸﺎن داده اﻧﺪ ﮐﻪ ﻣﻌﯿﺎر ﻫﺎي ﻣﺘﻌﺪدي ﺑﺮ روي ﺣﺠﻢ ﻧﻤﻮﻧﻪ ﺗﺎﺛﯿﺮ ﻣﯽ ﮔﺬارﻧﺪ ﮐﻪ ﺷﺎﻣﻞ اﻧﺪازه اﺛﺮ، ﺗـﻮان آزﻣﻮن، ﺳﻄﺢ ﻣﻌﻨﯽ داري، ﺣﺪاﮐﺜﺮ ﺧﻄﺎي اﻧﺪازه ﮔﯿﺮي، ﻧﺴﺒﺖ و ﭘﺮاﮐﻨﺪﮔﯽ ﺻﻔﺖ ﻣﻮرد ﺑﺮرﺳﯽ ﻣﯽ ﺑﺎﺷﻨﺪ. در اﯾﻦ ﻣﻄﺎﻟﻌﻪ ﻧﺸـﺎن داد ه ﺷﺪ ﮐﻪ در ﻣﺪل ﻫﺎي رﮔﺮﺳﯿﻮﻧﯽ ﺑﺎﯾﺴﺘﯽ اﺛﺮ ﺿﺮﯾﺐ ﻫﻤﺒﺴﺘﮕﯽ ﻣﺎﺑﯿﻦ ﻣﺘﻐﯿﺮﻫﺎ را ﻧﯿﺰ ﮐﻨﺘﺮل ﮐﺮده و ﺣﺠـﻢ ﻧﻤﻮﻧـﻪ را ﺑـﺎ در ﻧﻈـﺮ ﮔـﺮﻓﺘﻦ و ﺑﺎﻻﻧﺲ ﻣﻘﺪار ﭘﺮاﮐﻨﺪﮔﯽ اﯾﺠﺎد ﺷﺪه ﺑﺮآورد ﮐﺮد.ﻫﻤﭽﻨﯿﻦ اﻧﺪازه اﺛﺮ ﻣﻬﻤﺘﺮﯾﻦ ﺷﺎﺧﺺ و ﻧﯿﺰ ﺗﻮان آزﻣﻮن ﺑﻪ ﺷـﺪت ﺗﺤـﺖ ﺗـﺎﺛﯿﺮ ﺣﺠـﻢ ﻧﻤﻮﻧﻪ ﻣﯽ ﺑﺎﺷﺪ.
ﻧﺘﯿﺠﻪ ﮔﯿﺮي: در ﺑﯿﺸﺘﺮ ﻣﻄﺎﻟﻌﺎت ﻣﺸﺎﺑﻪ اﺳﺘﻨﺒﺎط ﺷﺪه اﺳﺖ ﮐﻪ ﺟﻬﺖ ﺑﺮآورد ﺣﺠﻢ ﻧﻤﻮﻧﻪ اي ﮐﺎرا و ﻣﻌﺘﺒﺮ ﺑﺎﯾﺴﺘﯽ ﻣﻘﺪار ﺧﻄﺎ ﭘـﺎﺋﯿﻦ و ﺗﻮان آزﻣﻮن ﺑﺎﻻ ﺑﻮده ﻫﻤﭽﻨﯿﻦ ﻣﻘﺎدﯾﺮ اﺛﺮ، ﺳﻄﺢ ﻣﻌﻨﯽ داري و ﭘﺮاﮐﻨﺪﮔﯽ و ﻫﻤﺒﺴﺘﮕﯽ ﺑﯿﻦ ﻣﺘﻐﯿﺮﻫﺎ ﮐﻨﺘﺮل ﺷﻮﻧﺪ.
چكيده لاتين :
One of the problems of the researchers at the start of the research is
estimation of sample size and selection of appropriate sample of the population. A small
sample size can damage research credibility and lead to loss of money and waste of time. In
contrast, a large sample size can lead to increased workload, cost, and lack of proper
feedback. The purpose of this study was to evaluate the sample size in various medical
studies.
Materials and Methods: In this study, after determining the types of studies, various sources
(books, journals, articles, etc.) were searched and different sampling methods based on the
type of variable, were determined and classified on the basis of the corresponding formulas.
Among the different methods we extracted the best formulas and methods on the basis of the
type of study and other important and confounding factors.
Result: Various studies have shown that several factors affect sample size including effect
size, power of the test, significance level, maximum error of measurement, ratio, and variation
of the trait. In this study, it was shown that in regression models, the effect of the correlation
coefficient between variables should be controlled and the sample size should be estimated by
considering and balancing the generated dispersion amount. Also, the size of the effect of the
most important indicator and the test power is strongly influenced by the sample size.
Conclusion: In most similar studies, it has been shown that for estimation of the effective and
valid sample size, the error rate should be low and the test power should be high, and also the
effect values, level of significance and dispersion, and the correlation between the variables
should be controlled.
