‫ﺍﻟﺘﺤﻠﻴﻞ ﺍﻟﺒﹸﻌﺪﻱ ﻭﺩﻭﺭﻩ ﰲ ﺍﲣﺎﺫ ﺍﻟﻘﺮﺍﺭ‬

‫د‪ .‬ﻣﺼﻄﻔﻰ ﺑﺎﺑﻜﺮ‬

 ‫ﻧﻈﺮﻳﺔ ﺑﻴﺰ‪ ،‬اﻟﺘﺤﻠﻴﻞ اﻟﺒﻌﺪي ودورﻩ ﻓﻲ اﺗﺨﺎذ اﻟﻘﺮار‬

‫اﻟﺘﺤﻠﻴﻞ اﻟﺒﻌﺪي )‪(Posterior Analysis‬‬

‫• ﻳﻌﺮﻑ ﺍﻟﺘﺤﻠﻴﻞ ﺍﻟﺬﻱ ﺗﺴﻨﺪ ﻓﻴﻪ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺴﺒﻘﺔ ﳊﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ‬
‫ﺑﺎﻟﺘﺤﻠﻴﻞ ﺍﻟﻘﺒﻠﻲ ﺃﻭ ﺍﳌﺴﺒﻖ )‪ (Prior Analysis‬ﻛﻤﺎ ﰲ ﺍﻷﺟﺰﺍﺀ‬
‫ﺍﻟﺴﺎﺑﻘﺔ‪ .‬ﺑﻨﺎﺀ‪ ‬ﻋﻠﻰ ﺍﻟﺘﺤﻠﻴﻞ ﺍﳌﺴﺒﻖ ﻋﻠﻴﻨﺎ ﺃﻥ �ﻘﺮﺭ ﰲ ﺃﻥ �ﺘﺨﺬ ﺍﻟﻔﻌﻞ‬
‫ﺍﻷﻣﺜﻞ ﺍﻟﺬﻱ ﻳﻘﱰﺣﻪ ﺍﻟﺘﺤﻠﻴﻞ ﺍﳌﺴﺒﻖ ﺃﻭ ﺍﻻ�ﺘﻈﺎﺭ ﻭﺍﻟﺒﺤﺚ ﻋﻦ ﻣﻌﻠﻮﻣﺎﺕ‬
‫ﺇﺿﺎﻓﻴﺔ ﻣﻦ ﺧﻼﻝ ﺍﻟﺒﺤﺚ ﺍﻟﺴﻮﻗﻲ‪ ،‬ﻣﺴﻮﺣﺎﺕ ﺍﳌﺴﺘﻬﻠﻚ‪ ،‬ﺩﺭﺍﺳﺔ ﳌﻌﺎﻳﻨﺔ‬
‫ﺍﳌﻨﺘﺞ ﺃﻭ ﺃﻱ ﻃﺮﻕ ﺃﺧﺮﻯ ﻭﺗﺴﻤﻰ ﻫﺬﻩ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﲟﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‬
‫)‪.(Sample Information‬‬
‫• �ﺴﺘﺨﺪﻡ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﰲ ﺗﻨﻘﻴﺢ ﻭﺗﺼﺤﻴﺢ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺴﺒﻘﺔ‬
‫ﳊﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ ﻭﺫﻟﻚ ﺑﺪﻣﺞ ﺍﻻﺛﻨﲔ ﻣﻌﺎً ﺑﺎﺳﺘﺨﺪﺍﻡ ﻗﺎ�ﻮﻥ ﺑﻴﺰ ﺍﻟﺬﻱ‬
‫ﺳﺒﻘﺖ ﺍﻹﺷﺎﺭﺓ ﺇﻟﻴﻪ ﻋﻨﺪ ﺍﺳﺘﻌﺮﺍﺽ ﺍﻻﺣﺘﻤﺎﻻﺕ ﻹﳚﺎﺩ‬
‫ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺼﺤﺤﺔ ﺃﻭ ﺍﳌﻨﻘﺤﺔ ﻭﺍﻟﱵ ﺗﻌﺮﻑ ﺑﺎﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ‬
‫)‪ .(Posterior Probabilities‬ﻭﻳﻌﺮﻑ ﺣﺴﺎﺏ ﺍﻟﻘﻴﻤﺔ‬
‫ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ‪ EMV‬ﺃﻭ ﺍﻟﺘﻜﻠﻔﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻟﻠﻔﺮﺹ ﺍﻟﻀﺎﺋﻌﺔ‬
‫‪ EOL‬ﺑﺎﺳﺘﺨﺪﺍﻡ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ﺑﺎﻟﺘﺤﻠﻴﻞ ﺍﻟﺒﻌﺪﻱ‪.‬‬
 ‫• ﺍﻓﱰﺽ ﺃﻥ ﻫﻨﺎﻟﻚ ﻋﺪﺩ ‪ k‬ﻣﻦ ﺍﳌﺆﺷﺮﺍﺕ ﳝﻜﻦ ﺃﻥ ﺗﺸﲑ ﳍﺎ ﺩﺭﺍﺳﺔ ﺍﻟﻌﻴﻨﺔ‬
‫‪ . I 1 , I 2 ,... I k‬ﺍﻓﱰﺽ ﺃ�ﻨﺎ ﺑﻌﺪ ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﻋﺮﻓﻨﺎ‬
‫ﺃﻱ ﺍﳌﺆﺷﺮﺍﺕ ‪ I j‬ﻗﺪ ﺣﺪﺙ ﻓﻌﻼ‪ ‬ﻭﻋﻠﻴﻪ �ﻮﺩ ﺃﻥ �ﺴﺘﺨﺪﻡ ﻫﺬﻩ ﺍﳌﻌﻠﻮﻣﺔ ﰲ‬
‫ﺗﺼﺤﻴﺢ ﺗﻘﺪﻳﺮ�ﺎ ﺍﳌﺴﺒﻖ ﻻﺣﺘﻤﺎﻻﺕ ﺣﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ ﺍﳌﺨﺘﻠﻔﺔ‪ .‬ﻭﺗﺆﺩﻱ‬
‫ﻫﺬﻩ ﺍﻟﻌﻤﻠﻴﺔ ﺇﱃ ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ﻭﺍﻟﱵ ﻳﻌﱪ ﻋﻨﻬﺎ ﺑﺼﻴﻐﺔ‬
‫ﺍﻻﺣﺘﻤﺎﻝ ﺍﳌﺸﺮﻭﻁ )‪ P(Si I j ) (Conditional Probability‬ﻭﺗﻘﺮﺃ‬
‫ﺍﺣﺘﻤﺎﻝ ﺍﳊﺎﻟﺔ ‪ Si‬ﺇﺫﺍ ﻋﺮﻓﻨﺎ ﺍﳌﻌﻠﻮﻣﺔ ﺃﻭ ﺍﳌﺆﺷﺮ ‪ I j‬ﻣﻦ ﺩﺭﺍﺳﺔ ﺍﻟﻌﻴﻨﺔ‪.‬‬
‫• ﻟﻜﻲ ﻳﺘﺴﻨﻰ ﻟﻨﺎ ﺣﺴﺎﺏ ﺍﻻﺣﺘﻤﺎﻝ ﺍﳌﺸﺮﻭﻁ ) ‪ P(Si I j‬ﳚﺐ ﻋﻠﻴﻨﺎ ﺃﻥ‬
‫�ﻌﺮﻑ ﺷﺊ ﻋﻦ ﺍﻟﻌﻼﻗﺔ ﺍﻻﺣﺘﻤﺎﻟﻴﺔ ﺑﲔ ﺍﳌﺆﺷﺮﺍﺕ ﻭﺣﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ‪ ،‬ﺃﻱ ﺃ�ﻨﺎ‬
‫ﳓﺘﺎﺝ ﳌﻌﺮﻓﺔ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺸﺮﻭﻃﺔ ) ‪ P(I j Si‬ﻟﻜﻞ ﺍﳌﺆﺷﺮﺍﺕ ﻭﻛﻞ‬
‫ﺣﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ‪ .‬ﻭﳝﻜﻦ ﺇﳚﺎﺩ ﻫﺬﻩ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻷﺧﲑﺓ ﻣﻦ ﺍﻟﺒﻴﺎ�ﺎﺕ‬
‫ﺍﻟﺘﺎﺭﳜﻴﺔ ﺃﻭ ﻣﻦ ﺍﻟﺘﻘﺪﻳﺮ ﺍﻟﺬﺍﺗﻲ ﰲ ﺣﺎﻟﺔ ﻋﺪﻡ ﺗﻮﻓﺮ ﺍﻟﺒﻴﺎ�ﺎﺕ ﺍﻟﺘﺎﺭﳜﻴﺔ‪.‬‬
‫ﻭﲤﺜﻞ ﻫﺬﻩ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺸﺮﻭﻃﺔ ﺇﻣﻜﺎ�ﻴﺔ ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﺍﳌﻌﻠﻮﻣﺔ ‪ I j‬ﻣﻦ‬
‫ﺩﺭﺍﺳﺔ ﺍﻟﻌﻴﻨﺔ ﰲ ﻭﺟﻮﺩ ﺣﺎﻟﺔ ﺍﻟﻄﺒﻴﻌﺔ ‪. Si‬‬
‫• ﺗﻌﺮﻑ ﻋﻤﻠﻴــﺔ ﲢﺪﻳﺪ ﻣﺎ ﺇﺫﺍ ﻛﺎ�ﺖ ﻫﻨﺎﻟﻚ ﺣﺎﺟﺔ ﳉﻤﻊ ﺍﻟﺒﻴﺎ�ﺎﺕ‬
‫ﺍﻹﺿﺎﻓﻴـــــﺔ ﻗﺒﻞ ﺍﲣــــــﺎﺫ ﺍﻟﻘﺮﺍﺭ ﺍﻟﻔﺎﺻﻞ ﺑﺎﻟﺘﺤﻠﻴﻞ ﻗﺒــﻞ ﺍﻟﺒﻌــــــــﺪﻱ‬
‫)‪ .(Preposterior Analysis‬ﻭﻟﺘﻮﺿﻴﺢ ﻫﺬﻩ ﺍﻟﻌﻤﻠﻴـﺔ ﻭﻛﻴﻔﻴﺔ‬
‫ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﻭﻭﺿﻌﻬﺎ ﰲ ﺻﻴﻐﺔ ﺍﻻﺣﺘﻤﺎﻻﺕ‬
‫ﺍﳌﺸﺮﻭﻃﺔ �ﻌﻄﻲ ﺍﳌﺜﺎﻝ ﺍﻟﺘﺎﱄ ) ‪. P(I j Si‬‬
 ‫ﻣﺜﺎﻝ ‪:1‬‬
‫ﺗﻮﺩ ﺷﺮﻛﺔ ﻣﺎﻛﺪﻭ�ﺎﻟﺪ ﺇ�ﺸﺎﺀ ﻣﺒﻨﻰ ﺟﺪﻳﺪ ﺑﺎﻟﻘﺮﺏ ﻣﻦ ﻣﺮﻛﺰ ﻟﻠﺘﺴﻮﻕ ﻳﺆﻣﻪ‬
‫ﺍﻟﺴﻴﺎﺡ‪ .‬ﻫﻨﺎﻟﻚ ﺛﻼﺛﺔ ﺃﺣﺠﺎﻡ ﳑﻜﻨﺔ ﻟﻠﻤﺒﻨﻰ ‪، (a )5000 ، (a )3000‬‬
‫‪2‬‬ ‫‪1‬‬

‫ﺃﻭ ‪ (a )9000‬ﻗﺪﻡ ﻣﺮﺑﻊ ﺍﻋﺘﻤﺎﺩﺍً ﻋﻠﻰ ﻣﺴﺘﻮﻯ ﺍﻟﻄﻠﺐ ﺍﻟﻴﻮﻣﻲ ﺍﳌﺘﻮﻗﻊ‬

‫‪3‬‬

‫ﺿﻌﻴﻒ ) ‪ ، ( s1‬ﻣﺘﻮﺳﻂ ﺃﻭ ﻣﺮﺗﻔﻊ‪ .‬ﻭﻗﺪ ﻗﺮﺭ ﻣﺴﺘﺸﺎﺭﻱ ﺍﻟﺸﺮﻛﺔ‬

‫ﺍﺣﺘﻤﺎﻻﺕ ﺍﻟﻄﻠﺐ ﺍﻟﻀﻌﻴﻒ ) ‪ ، ( s1‬ﺍﻟﻄﻠﺐ ﺍﳌﺘﻮﺳﻂ ) ‪ ( s2‬ﻭﺍﻟﻄﻠﺐ‬
‫ﺍﳌﺮﺗﻔﻊ ) ‪ ( s3‬ﺃﳖﺎ ﺗﺴﺎﻭﻱ ‪ ،0.3 ،0.2‬ﻭ ‪ 0.5‬ﻋﻠﻰ ﺍﻟﺘﻮﺍﱄ‪ .‬ﻭﻳﻮﺿﺢ‬
‫ﺍﳉﺪﻭﻝ ﺃﺩ�ﺎﻩ ﺍﳌﻨﺎﻓﻊ ﻭﺍﻟﻘﻴﻢ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻟﻸﻓﻌﺎﻝ‪ .‬ﺍﳌﻄﻠﻮﺏ ﲢﻠﻴﻞ‬
‫ﺍﻷﻓﻌﺎﻝ ﺍﳌﺘﺎﺣﺔ ﻟﻠﺸﺮﻛﺔ ﻓﻴﻤﺎ ﳜﺘﺺ ﺑﺎﳌﺸﺮﻭﻉ‪.‬‬
 ‫ﺍﳊﻞ‪:‬‬
‫• ﺑﻔﺤﺺ ﺍﻟﻌﻤﻮﺩ ﺍﻷﺧﲑ ﰲ ﺍﳉﺪﻭﻝ ﻳﺸﲑ ﺍﻟﺘﺤﻠﻴﻞ ﺍﻟﻘﺒﻠﻲ ﺃﻭ ﺍﳌﺴﺒﻖ ﺃﻥ‬
‫ﺇ�ﺸﺎﺀ ﻣﺒﻨﻰ ﻛﺒﲑ ﺍﳊﺠﻢ )‪ 9000‬ﻗﺪﻡ ﻣﺮﺑﻊ( ﻫﻮ ﺍﻟﻘﺮﺍﺭ ﺍﻷﻣﺜﻞ ﺣﻴﺚ‬
‫ﳛﻘﻖ ﺃﻋﻠﻰ ﺭﺑﺢ ﻣﺘﻮﻗﻊ ﻟﻠﺸﺮﻛﺔ ﺑﲔ ﻛﻞ ﺍﳋﻴﺎﺭﺍﺕ ﺍﳌﺘﺎﺣﺔ ‪.($24,000‬‬
 ‫ﰲ ﺍﳌﻘﺎﺑﻞ ﺇﺫﺍ ﻋﺮﻓﻨﺎ ﻣﺴﺒﻘﺎً ﺃﻥ ﺣﺎﻟﺔ ﺍﻟﻄﻠﺐ ﻫﻲ ‪ s1‬ﻓﺎﳋﻴﺎﺭ ﺍﻷﻣﺜﻞ ﻫﻮ‬ ‫•‬
‫ﻟﻄﻠﺐ‬
‫‪ ، a1‬ﺣﺎﻟﺔ ﺍﻟﻄﻠﺐ ‪ s2‬ﻓﺎﳋﻴﺎﺭ ﺍﻷﻣﺜﻞ ﻫﻮ ‪ ، a2‬ﻭﺣﺎﻟﺔ ﺍ ‪s3‬‬
‫ﻓﺎﳋﻴﺎﺭ ﺍﻷﻣﺜﻞ ﻫﻮ ‪ a3‬ﻭﻋﻠﻴﻪ ﰲ ﺣﺎﻟﺔ ﺍﳌﻌﺮﻓﺔ ﺍﻟﺘﺎﻣﺔ ﺗﻜﻮﻥ ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ‬
‫‪ . (10,000×0.2 + 25,000×0.3+ 60,000×0.5) $39,000‬ﻭﻋﻠﻴﻪ ﻓﺈﻥ‬
‫ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻟﻠﻤﻌﺮﻓﺔ ﺍﻟﺘﺎﻣﺔ ﺗﺴﺎﻭﻱ ‪.(39,000-24,000) $15,000‬‬
‫ﺃﻱ ﺃﻥ ﻋﻠﻰ ﺍﻟﺸﺮﻛﺔ ﺃﻻ ﺗﻨﻔﻖ ﺃﻛﺜﺮ ﻣﻦ ‪ $15,000‬ﻟﻠﺤﺼﻮﻝ ﻋﻠﻰ ﺃﻱ‬
‫ﻣﻌﻠﻮﻣﺎﺕ ﺇﺿﺎﻓﻴﺔ ﺗﺴﺎﻋﺪ ﰲ ﺍﲣﺎﺫ ﺍﻟﻘﺮﺍﺭ ﺑﺸﺄﻥ ﺍﳌﺸﺮﻭﻉ‪.‬‬
‫ﺍﻓﱰﺽ ﺃﻥ ﺍﻟﺸﺮﻛﺔ ﺍﺳﺘﺄﺟﺮﺕ ﺑﻴﺖ ﻭﻟﺴﻦ ﻟﻼﺳﺘﺸﺎﺭﺍﺕ ﻹﺟﺮﺍﺀ‬ ‫•‬
‫ﻣﺴﺢ ﺳﻮﻗﻲ ﻟﻠﺤﺼﻮﻝ ﻋﻠﻰ ﻣﻌﻠﻮﻣﺎﺕ ﺇﺿﺎﻓﻴﺔ ﻋﻦ ﻣﺴﺘﻮﻯ ﺍﻟﻄﻠﺐ‪.‬‬
‫ﻭﻗﺪ ﺃﺷﺎﺭ ﺗﻘﺮﻳﺮ ﺑﻴﺖ ﺍﻻﺳﺘﺸﺎﺭﺍﺕ ﺃﻥ ﻣﺴﺘﻮﻯ ﺍﻟﻄﻠﺐ ﺳﻴﻜﻮﻥ‬
‫ﻣﺘﻮﺳﻂ‪.‬‬
‫ﺇﺫﺍ ﺍﻋﺘﻤﺪﺕ ﺍﻟﺸﺮﻛﺔ ﻋﻠﻰ ﺍﻟﺘﻘﺮﻳﺮ ﺑﺄ�ﻪ ‪ %100‬ﺻﺤﻴﺢ ﺗﻜﻮﻥ‬ ‫•‬
‫ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺼﺤﺤﺔ ﻫﻲ ‪ P ( s2 ) = 1 ، P ( s1 ) = 0‬ﻭ ‪P ( s3 ) = 0‬‬
‫ﻭﻳﻜﻮﻥ ﺍﻟﻔﻌﻞ ﺍﻷﻣﺜﻞ ﻫﻮ ‪ .a2‬ﻏﲑ ﺃ�ﻪ ﻣﻦ ﺍﳋﻄﺄ ﺍﻋﺘﺒﺎﺭ ﺍﻟﺘﻘﺮﻳﺮ ‪%100‬‬
‫ﺻﺤﻴﺢ ﻭﺑﺎﻟﺘﺎﱄ ﻣﺴﺎﻭﺍﺓ ﺍﳌﺆﺷﺮﺍﺕ ﻣﻦ ﺩﺭﺍﺳﺔ ﺍﻟﻌﻴﻨﺔ ﲝﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ‬
‫ﻷﻥ ﺩﺭﺍﺳﺔ ﺍﻟﻌﻴﻨﺔ ﻣﻌﺮﺿﺔ ﻷﺧﻄﺎﺀ ﻋﺪﻳﺪﺓ ﺑﺎﻹﺿﺎﻓﺔ ﳋﻄﺄ ﺍﳌﻌﺎﻳﻨﺔ‪.‬‬
‫ﻋﻠﻴﻪ ﻻ ﺑﺪ ﻣﻦ ﺗﻘﻴﻴﻢ ﺑﻴﺎ�ﺎﺕ ﺍﻟﻌﻴﻨﺔ ﺑﺎﻟﺮﺟﻮﻉ ﺇﱃ ﺍﻟﺴﺠﻞ ﺍﻟﺘﺎﺭﳜﻲ‪.‬‬
‫ﺍﻓﱰﺽ ﺃﻥ ﻫﻨﺎﻟﻚ ﺛﻼﺙ ﻣﺆﺷﺮﺍﺕ ﳝﻜﻦ ﺍﳊﺼﻮﻝ ﻋﻠﻴﻬﺎ ﻣﻦ‬ ‫•‬
‫ﺑﻴﺎ�ﺎﺕ ﺍﻟﻌﻴﻨﺔ‪ I1 :‬ﻳﺸﲑ ﺇﱃ ﺿﻌﻒ ﺍﻟﻄﻠﺐ‪ I 2 ،‬ﻳﺸﲑ ﺇﱃ‬
‫ﻛﻮﻥ ﺍﻟﻄﻠﺐ ﻣﺘﻮﺳﻂ‪ ،‬ﻭ ‪ I 3‬ﻳﺸﲑ ﺇﱃ ﺍﺭﺗﻔﺎﻉ ﺍﻟﻄﻠﺐ‪.‬‬
‫ﺍﻓﱰﺽ ﺃ�ﻪ ﺑﻨﺎﺀ‪ ‬ﻋﻠﻰ ﺍﳌﺴﻮﺣﺎﺕ ﺍﻟﺴﺎﺑﻘﺔ ﺃﻥ ﺍﺣﺘﻤﺎﻻﺕ‬
‫ﺻﺤﺔ ﺗﻘﺪﻳﺮﺍﺕ ﺑﻴﺖ ﺍﻻﺳﺘﺸﺎﺭﺍﺕ ﻟﻠﻤﺆﺷﺮﺍﺕ ﰲ ﻇﻞ‬
‫ﺣﺎﻻﺕ ﺍﻟﻄﺒﻴﻌﺔ ﺍﻟﺜﻼﺛﺔ ⎠⎟⎞ ‪ ⎛⎜⎝ P ⎛⎜⎝ I j Si ⎞⎟⎠ , i = 1,2,3‬ﻫﻲ ﻛﻤﺎ‬
‫ﻳﺒﻴﻨﻬﺎ ﺍﳉﺪﻭﻝ ﺃﺩ�ﺎﻩ‪:‬‬
‫ﳊﺴﺎﺏ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ) ‪ P(Si I j‬ﻣﻦ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺴﺒﻘﺔ‬ ‫•‬
‫) ‪ P( si‬ﻭﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ) ‪� P (I j Si‬ﺴﺘﺨﺪﻡ �ﻈﺮﻳﺔ ﺑﻴﺰ‪.‬‬
 ‫�ﻈﺮﻳﺔ ﺑﻴﺰ‬
‫د‪ .‬ﻣﺼﻄﻔﻰ ﺑﺎﺑﻜﺮ‬
 ‫ﻧﻈﺮﻳﺔ ﺑﻴﺰ )‪(Bayes’ Theorem‬‬
‫• ﻳﻌﺘﱪ ﺑﻴﺰ ﻣﻦ ﺃﻭﺍﺋﻞ ﺍﻹﺣﺼﺎﺋﻴﲔ ﺍﻟﺬﻳﻦ ﺃﺷﺎﺭﻭﺍ ﺇﱃ ﺇﻣﻜﺎ�ﻴﺔ ﺗﻨﻘﻴﺢ ﺃﻭ‬
‫ﺗﺼﺤﻴﺢ ﺍﻻﺣﺘﻤﺎﻻﺕ ﰲ ﺿﻮﺀ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﺍﻟﻼﺣﻘﺔ ﻭﻣﺴﺎﳘﺔ ﺑﻴﺰ ﰲ‬
‫ﻫﺬﺍ ﺍ‪‬ﺎﻝ ﻫﻮ ﺇﻋﻄﺎﺀ ﻃﺮﻳﻘﺔ ﳊﺴﺎﺏ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺸﺮﻭﻃﺔ‪.‬‬
‫• ﻟﺸﺮﺡ �ﻈﺮﻳﺔ ﺑﻴﺰ ﺍﻓﱰﺽ ﺃﻥ ﳎﺎﻝ ﺍﻟﻌﻴﻨﺔ ﻟﺘﺠﺮﺑﺔ ﻋﺸﻮﺍﺋﻴﺔ ﻫﻮ ﺍﻷﺣﺪﺍﺙ‬
‫ﺍﳌﺘﻨﺎﻓﻴﺔ ‪ Ek ...، E2 ، E1‬ﺫﺍﺕ ﺍﺣﺘﻤﺎﻻﺕ ﺍﳊﺪﻭﺙ) ‪، P( E2 )، P( E1‬‬
‫‪ P( Ek )...‬ﻭﺍﻟﱵ ﺗﻌﺮﻑ ﺑﺎﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺴﺒﻘﺔ ﻷﳖﺎ ﲢﺪﺩ ﻗﺒﻞ ﺃﺧﺬ‬
‫ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﰲ ﺍﳊﺴﺒﺎﻥ‪ .‬ﻛﻤﺎ ﺗﺴﻤﻰ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﻨﻘﺤﺔ ﰲ‬
‫ﺿﻮﺀ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﺑﺎﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ﻷﳖﺎ ﲢﺴﺐ ﺑﻌﺪ ﺃﺧﺬ‬
‫ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﰲ ﺍﳊﺴﺒﺎﻥ‪.‬‬
‫• ﺇﺫﺍ ﻛﺎ�ﺖ ‪ A‬ﲤﺜﻞ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﻓﺈﻥ ﺍﺣﺘﻤﺎﻝ ﺍﳊﺪﺙ ‪ Ei‬ﻣﻌﻄﻲ‬
‫ﺍﳌﻌﻠﻮﻣﺔ ‪ A‬ﻳﻌﻄﻰ ﺑﺎﻟﻘﺎ�ﻮﻥ‪:‬‬
‫)‪P(Ei , A‬‬
‫= )‪P(Ei A‬‬
‫)‪P ( A‬‬

‫ﻭﻣﻦ ﺍﻟﻘﺎ�ﻮﻥ ﺍﻟﻌﺎﻡ ﻟﻀﺮﺏ ﺍﻻﺣﺘﻤﺎﻻﺕ‪:‬‬

‫) ‪P ( A, E i ) = P ( E i ) P ( A E i‬‬
‫ﻭﺑﺘﻌﻮﻳﺾ �ﺘﻴﺠﺔ ﻗﺎ�ﻮﻥ ﺍﻟﻀﺮﺏ ﰲ ﺍﳌﻌﺎﺩﻟﺔ ﺍﻟﺴﺎﺑﻘﺔ ﳓﺼﻞ ﻋﻠﻰ ﺍﻻﺣﺘﻤﺎﻻﺕ‬
‫ﺍﻟﺒﻌﺪﻳﺔ ﻟﻠﺤﺪﺙ ‪: Ei‬‬
‫⎟⎞ ‪P ( E i ) P ⎛⎜ A E i‬‬
‫= ⎟⎞ ‪P ⎛⎜ E i A‬‬ ‫⎝‬ ‫⎠‬
‫⎝‬ ‫⎠‬ ‫)‪P ( A‬‬
‫ﻭﻳﻌﺮﻑ ﻫﺬﺍ ﺍﻟﻘﺎ�ﻮﻥ ﺑﻘﺎ�ﻮﻥ ﺑﻴﺰ‪.‬‬
‫• ﰲ ﻛﺜﲑ ﻣﻦ ﺗﻄﺒﻴﻘﺎﺕ �ﻈﺮﻳﺔ ﺑﻴــﺰ ﻻ ﻳﻌﻄﻰ ﺍﻻﺣﺘﻤﺎﻝ )‪ P( A‬ﻭﺍﻟﺬﻱ‬
‫ﻳﻌﺮﻑ ﺃﻳﻀﺎً ﲟﻘﺪﺍﺭ ﺍﳌﻌﻠﻮﻣــﺎﺕ )‪(Amount of Information‬‬
‫ﻣﺒﺎﺷﺮﺓ ﻭﻟﻜﻦ ﳝﻜﻦ ﺣﺴﺎﺑﻪ ﺑﺎﻟﻘﺎ�ﻮﻥ‪:‬‬

‫) ‪P(A) = P(A, E1) + P(A, E2 ) +...+ P(A, Ek‬‬

‫ﻭﺑﺘﻄﺒﻴﻖ ﻗﺎ�ﻮﻥ ﺍﻟﻀﺮﺏ‪:‬‬

‫(‬ ‫)‬ ‫(‬ ‫)‬ ‫(‬

‫‪P( A) = P(E ) P A E + P(E ) P A E + ...+ P(E ) P A E‬‬
‫‪1‬‬ ‫‪1‬‬ ‫‪2‬‬ ‫‪2‬‬ ‫‪k‬‬ ‫‪k‬‬
‫)‬
‫‪k‬‬
‫(‬
‫‪= ∑ P(Ei ) P A Ei‬‬ ‫)‬
‫‪i=1‬‬
 ‫ﻭﻋﻠﻴﻪ ﰲ ﻫﺬﻩ ﺍﳊﺎﻻﺕ ﻳﺼﺒﺢ ﻗﺎ�ﻮﻥ ﺑﻴﺰ‪:‬‬

‫⎟⎞ ‪P( Ei ) P⎛⎜ A Ei‬‬

‫= ⎟⎞‪P⎛⎜ Ei A‬‬ ‫⎝‬ ‫⎠‬
‫⎝‬ ‫⎠‬ ‫‪k‬‬ ‫⎛‬ ‫⎞‬
‫∑‬ ‫‪P‬‬ ‫(‬ ‫‪E‬‬ ‫‪j‬‬ ‫)‬ ‫‪P‬‬ ‫⎜‬ ‫‪A‬‬ ‫‪E‬‬ ‫‪j‬‬ ‫⎟‬
‫‪j =1‬‬ ‫⎝‬ ‫⎠‬
 ‫• ﻭﺑﺘﻄﺒﻴﻖ �ﻈﺮﻳﺔ ﺑﻴﺰ ﳊﺴﺎﺏ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ﻣﺘﺎﺑﻌﺔ ﳊﻞ ﺍﳌﺜﺎﻝ ‪1‬‬
‫ﳓﺼﻞ ﻋﻠﻰ ﺍﻟﻘﺎ�ﻮﻥ ﺍﻟﺘﺎﱄ‪:‬‬

‫⎟⎞ ‪P ( si ) P ⎛⎜ I j Si‬‬
‫= ⎟⎞ ‪P ⎛⎜ Si I j‬‬ ‫⎝‬ ‫⎠‬
‫⎝‬ ‫⎠‬ ‫⎛‬ ‫⎞‬
‫⎟ ‪∑ P ( sk ) P ⎜ I j S k‬‬
‫‪k‬‬ ‫⎝‬ ‫⎠‬

‫) ‪P ( I j ) = ∑ P ( sk ) P (I j S k‬‬ ‫ﺣﻴﺚ ‪ i ، j ، k = 1,2,3‬ﻭﺣﻴﺚ‬

‫ﻭﺍﳉﺪﻭﻝ ﺃﺩ�ﺎﻩ ﻳﻌﺮﺽ ﻗﻴﻢ ﻣﻜﻮ�ﺎﺕ ﺍﻟﻘﺎ�ﻮﻥ ﻟﻠﻤﺜﺎﻝ‪:‬‬
‫‪k‬‬
‫ﻭﺑﺘﻌﻮﻳﺾ ﻗﻴﻢ ﻣﻜﻮ�ﺎﺕ ﺍﻟﻘﺎ�ﻮﻥ ﳓﺼﻞ ﻋﻠﻰ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ‬
‫ﺍﳌﻮﺿﺤﺔ ﰲ ﺍﳉﺪﻭﻝ ﺍﻟﺘﺎﱄ‪:‬‬
‫• ﺑﻌﺪ ﻣﻌﺮﻓﺔ ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﻟﺒﻌﺪﻳﺔ ﳝﻜﻦ ﺣﺴﺎﺏ ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ‬
‫ﺍﻟﺒﻌﺪﻳﺔ )‪ (Posterior Expected Payoff‬ﺃﻭ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ‬
‫ﺍﳌﺘﻮﻗﻌﺔ ﻟﻠﻔﻌﻞ ‪ ai‬ﻣﻌﻄﻰ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ‪ I j‬ﺑﺎﻟﻘﺎ�ﻮﻥ‪:‬‬

‫‪K‬‬
‫⎟⎞ ‪EMV ⎛⎜ ai I j ⎞⎟ = ∑ π P⎛⎜ S I j‬‬
‫⎝‬ ‫‪⎠ k =1 ik ⎝ k‬‬ ‫⎠‬

‫ﻭﺑﺘﻄﺒﻴﻖ ﺍﻟﻘﺎ�ﻮﻥ ﻋﻠﻰ ﻣﺜﺎﻝ ﻣﺸﺮﻭﻉ ﺷﺮﻛﺔ ﻣﺎﻛﺪﻭ�ﺎﻟﺪ ﳓﺼﻞ ﻋﻠﻰ‬

‫ﺍﳉﺪﻭﻝ ﺍﻟﺘﺎﱄ ﻟﻠﻘﻴﻢ ﺍﳌﺘﻮﻗﻌﺔ ﺍﻟﺒﻌﺪﻳﺔ‪:‬‬
 ‫• ﺇﳚﺎﺩ ﺍﻟﻔﻌﻞ ﺍﻷﻣﺜﻞ‪:‬‬

‫ﻓﺈﻥ‬ ‫‪Ij‬‬‫∗ ﺑﺎﺗﺒﺎﻉ ﻣﻌﻴﺎﺭ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻭﲟﻌﺮﻓﺔ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‬
‫ﺍﻟﻔﻌﻞ ﺍﻷﻣﺜﻞ ﻫﻮ ﺍﻟﺬﻱ ﳛﻘﻖ ﺃﻋﻠﻰ ﻗﻴﻤﺔ ﻣﺘﻮﻗﻌﺔ ﺑﻌﺪﻳﺔ‪.‬‬
‫∗ ﺑﺘﻄﺒﻴﻖ ﺍﳌﻌﻴﺎﺭ ﻋﻠﻰ ﺟﺪﻭﻝ ﺍﻟﻘﻴﻢ ﺍﳌﺘﻮﻗﻌﺔ ﺍﻟﺒﻌﺪﻳﺔ ﻟﻠﻤﺜﺎﻝ ﺍﻟﺴﺎﺑﻖ ﳒﺪ‬
‫ﺃﻥ ﺍﻟﻔﻌﻞ ﺍﻷﻣﺜﻞ ﻫﻮ ‪ a1‬ﻋﻨﺪﻣﺎ ﺗﻜﻮﻥ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ‪ a2 ، I1‬ﻋﻨﺪﻣﺎ‬
‫ﺗﻜﻮﻥ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ‪ ، I 2‬ﻭ ‪ a3‬ﻋﻨﺪﻣﺎ ﺗﻜﻮﻥ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ‪. I 3‬‬
‫∗ ﺗﻮﺟﺪ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻟﻠﻘﺮﺍﺭ ﺍﻷﻣﺜﻞ ﰲ ﻇﻞ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‬
‫ﺑﻀﺮﺏ ﺍﻟﻘﻴﻢ ﺍﻟﻨﻘﺪﻳـﺔ ﺍﳌﺘﻮﻗﻌـﺔ ﻟﻸﻓﻌـﺎﻝ ﺍﳌﺜﻠﻰ ﺃﻋﻼﻩ ﰲ ﺍﺣﺘﻤﺎﻻﺕ‬
‫ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨـﺔ ﺍﳌﻘﺎﺑﻠــﺔ ﳍﺎ ﻭﻳﺮﻣﺰ ﳍﺬﻩ ﺍﻟﻘﻴﻤﺔ ﺑـــ ‪EPSI‬‬
‫ﺃﻭ )‪ ،(Expected Payoff with Sample Information‬ﺃﻱ‪:‬‬
‫‪k‬‬ ‫⎡‬ ‫⎤‬
‫) ‪EPSI = ∑ maxi ⎢EMV⎛⎜ ai I j ⎞⎟⎥ P(I j‬‬
‫‪j=1‬‬ ‫⎣‬ ‫⎝‬ ‫⎦⎠‬

‫ﻭﺑﺘﻄﺒﻴﻖ ﻗﺎ�ﻮﻥ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﰲ ﻇﻞ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﻋﻠﻰ ﻣﺜﺎﻝ‬

‫ﺷﺮﻛﺔ ﻣﺎﻛﺪﻭ�ﺎﻟﺪ ﳓﺼﻞ ﻋﻠﻰ ‪$27,688.65‬‬
‫)ﺃﻱ ‪( 13,200 × 0.25 + 24,530 × 0.42 + 42685 × 0.33‬‬
‫∗ ﻟﺘﻘﻴﻴﻢ ﺟﺪﻭﻯ ﲨﻊ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﻻ ﺑﺪ ﻣﻦ ﻣﻘﺎﺭ�ﺔ ﺗﻜﻠﻔﺔ‬
‫ﺍﳊﺼﻮﻝ ﻋﻠﻰ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺑﺎﻟﻔﺎﺋﺪﺓ ﺍﳌﺘﻮﻗﻌﺔ ﻣﻦ ﺍﺳﺘﺨﺪﺍﻡ ﺍﳌﻌﻠﻮﻣﺎﺕ‪.‬‬
‫ﻭﺗﻘــﺎﺱ ﺍﻟﻔﺎﺋــﺪﺓ ﺍﳌﺘﻮﻗﻌــﺔ ﲝﺴﺎﺏ ﺍﻟﻘﻴﻤــﺔ ﺍﳌﺘﻮﻗﻌﺔ ﳌﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‬
‫ﺃﻭ )‪(Expected Value of Sample Information‬‬
‫”‪“EVSI‬ﻭﺍﻟﱵ ﺗﻌﺮﻑ ﺑﺄﺧﺬ ﺍﻟﻔﺮﻕ ﺑﲔ ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ ﺍﳌﺜﻠﻰ ﰲ ﻇﻞ‬
‫ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﻭﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ ﺍﳌﺜﻠﻰ ﰲ ﻏﻴﺎﺏ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‪،‬‬
‫ﺃﻱ‪:‬‬
‫∗ ‪EVSI = EPSI − EMV‬‬

‫ﺣﻴﺚ ∗ ‪ EMV‬ﻫﻲ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﻟﻠﻔﻌﻞ ﺍﻷﻣﺜﻞ ﺑﺎﺳﺘﺨﺪﺍﻡ‬

‫ﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺴﺒﻘﺔ ﻭﰲ ﻋﺪﻡ ﻭﺟﻮﺩ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ‪.‬‬
‫∗ ﺑﺎﻟﺘﻄﺒﻴﻖ ﻋﻠﻰ ﻣﺜﺎﻝ ﺷﺮﻛﺔ ﻣﺎﻛﺪﻭ�ﺎﻟﺪ ﳒﺪ ﺃﻥ ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ ﳌﻌﻠﻮﻣﺎﺕ‬
‫ﺍﻟﻌﻴﻨﺔ ﺗﺴﺎﻭﻱ ‪) 3,688.5‬ﺃﻱ ‪ .(27,688-24,000‬ﻋﻠﻴﻪ ﻓﻬﻨﺎﻟﻚ ﺟﺪﻭﻯ‬
‫ﳉﻤﻊ ﺍﳌﻌﻠﻮﻣﺎﺕ ﺍﻹﺿﺎﻓﻴﺔ ﺇﺫﺍ ﻛﺎ�ﺖ ﺗﻜﻠﻔﺘﻬﺎ ﺃﻗﻞ ﻣﻦ ﺃﻭ ﺗﺴﺎﻭﻱ‬
‫‪.$3,688.65‬‬
‫∗ ﻳﻌﺮﻑ ﺍﻟﻔﺮﻕ ﺑﲔ ﺍﻟﻘﻴﻤﺔ ﺍﳌﺘﻮﻗﻌﺔ ﳌﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﻭﺗﻜﻠﻔﺔ ﺍﻟﻌﻴﻨﺔ ﺑﺼﺎﰲ‬
‫ﺍﳌﻜﺎﺳﺐ ﺍﳌﺘﻮﻗﻌﺔ ﻣﻦ ﺍﺳﺘﺨﺪﺍﻡ ﺍﻟﻌﻴﻨﺔ ) ‪Expected Net Gain‬‬
‫‪ (From Sampling‬ﻣﺜﻼ‪ ‬ﺇﺫﺍ ﻭﺍﻓﻖ ﺑﻴﺖ ﻭﻟﺴﻦ ﻟﻼﺳﺘﺸﺎﺭﺍﺕ ﻋﻠﻰ‬
‫ﺗﻮﻓﲑ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﺑﺘﻜﻠﻔﺔ ‪ $2000‬ﻳﻜﻮﻥ ﺻﺎﰲ ﺍﳌﻜﺎﺳﺐ ﺍﳌﺘﻮﻗﻌﺔ ﻣﻦ‬
‫ﺍﺳﺘﺨﺪﺍﻡ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﰲ ﺍﳌﺜﺎﻝ ﻣﺴﺎﻭﻳﺎً ﻟـــ ‪.$1,688.65‬‬
‫∗ ﺃﺧﲑﺍً ﺇﺫﺍ ﺃﺷﺎﺭ ﻛﻤﺎ ﺫﻛﺮ�ﺎ ﺗﻘﺮﻳﺮ ﺑﻴﺖ ﺍﳋﱪﺓ ﺃﻥ ﻣﺆﺷﺮ ﺍﻟﻌﻴﻨﺔ‬
‫ﻫﻮ ‪ I 2‬ﻓﺈﻥ ﺍﻟﻘﺮﺍﺭ ﺍﻷﻣﺜﻞ ﻫﻮ ﺍﲣﺎﺫ ﺍﻟﻔﻌﻞ ‪ a2‬ﺃﻭ ﺗﺸﻴﻴﺪ‬
‫ﻣﺒﻨﻰ ﻣﺘﻮﺳﻂ ﺍﳊﺠﻢ ﺣﻴﺚ ﺍﻟﻘﻴﻤﺔ ﺍﻟﻨﻘﺪﻳﺔ ﺍﳌﺘﻮﻗﻌﺔ ﰲ ﻇﻞ‬
‫ﺍﳌﻌﻠﻮﻣﺔ ‪ I 2‬ﺃﻋﻠﻰ ﻣﺎ ﺗﻜﻮﻥ ﻋﻨﺪ ﺍﲣﺎﺫ ﺍﻟﻔﻌﻞ ‪. a2‬‬
‫∗ ﰲ ﺍﳋﺘﺎﻡ �ﻜﺮﺭ ﺍﻟﺘﻨﻮﻳﻪ ﺃﻥ ﺻﺤﺔ ﺍﻟﺘﺤﻠﻴﻼﺕ ﰲ ﻫﺬﺍ ﺍﳉﺰﺀ‬
‫ﺗﻌﺘﻤﺪ ﻛﺜﲑﺍً ﻋﻠﻰ ﺩﻗﺔ ﺗﻘﺪﻳﺮ�ﺎ ﳉﺪﻭﻝ ﺍﳌﻨﺎﻓﻊ‪ ،‬ﺍﻻﺣﺘﻤﺎﻻﺕ‬
‫ﺍﳌﺴﺒﻘﺔ ﻭﺍﻻﺣﺘﻤﺎﻻﺕ ﺍﳌﺒﻨﻴﺔ ﻋﻠﻰ ﻣﻌﻠﻮﻣﺎﺕ ﺍﻟﻌﻴﻨﺔ ﻭﻟﺬﻟﻚ‬
‫ﳚﺐ ﺇﺟﺮﺍﺀ ﺍﺧﺘﺒﺎﺭﺍﺕ ﺍﻟﺘﺄﺛﺮ )‪(Sensitivity Analysis‬‬
‫ﺣﻮﻝ ﻫﺬﻩ ﺍﻟﻘﻴﻢ ﻭﲢﺪﻳﺪ ﻣﺪﻯ ﺣﺴﺎﺳﻴﺔ ﺍﻟﻘﺮﺍﺭ ﺍﻷﻣﺜﻞ‬
‫ﻟﻠﺘﻐﲑﺍﺕ ﺍﻟﻄﻔﻴﻔﺔ ﰲ ﻫﺬﻩ ﺍﻟﻘﻴﻢ‪.‬‬