COMMunication
RESearch.NET

User Tools

Site Tools

You are here: index » sequential_regression
Trace: • introduction_to_data_science • examples_r • assignment_example • c • sequential_regression • multiple_program_provider • probability • answer_ex2 • descriptive_stat • multiple_regression_lecture_note_for_r
sequential_regression

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
sequential_regression [2020/12/01 15:04] – [Report] hkimscilsequential_regression [2024/06/12 08:30] (current) – [r] hkimscil
Line 1: Line 1:
 +====== Sequential or Hierarchical regression ======
 +연구자가 판단하여 독립변인들 중 필요한 것들을 묶어서 스테이지 별로 (단계 별) 넣고 분석하는 것을 말한다. Stepwise regression은 이를 컴퓨터나 계산방법을 통하여 수행하게 된다.  
 ====== 데이터 ====== ====== 데이터 ======
 ^  DATA for regression analysis   ^^^ ^  DATA for regression analysis   ^^^
Line 44: Line 46:
 The below is just an exercise for figuring out the unique part of r<sup>2</sup> value for x1 and x2 (수입, 가족수). For more information see part and zero-order relationship: see [[:multiple_regression#determining_ivs_role]] in multiple regression The below is just an exercise for figuring out the unique part of r<sup>2</sup> value for x1 and x2 (수입, 가족수). For more information see part and zero-order relationship: see [[:multiple_regression#determining_ivs_role]] in multiple regression
  
-|  zero-order  ||  part  ||+|  zero-order  ||  part = semi-partial  ||
 | x1  | x2  | x1p  | x2p  | | x1  | x2  | x1p  | x2p  |
 | .794  | -.692  | .565  | -.409  |  | .794  | -.692  | .565  | -.409  | 
 |  zero-order square  ||  part (in spss) = semipartial (in general)  || |  zero-order square  ||  part (in spss) = semipartial (in general)  ||
-| x1 sq (x1sq)  | x2 sq (x1sq)  | x1 part sq (x1psq) | x2 part sq (x1psq)  |+| x1 zsq (x1zsq)  | x2 zsq (x1zsq)  | x1 semi-partial (or part) sq (x1spsq) | x2 part sq (x1spsq)  |
 | .630436  | .478864  | .319225  | .167281  | | .630436  | .478864  | .319225  | .167281  |
 | a+b / a+b+c+d  | b+c / a+b+c+d  | a / a+b+c+d  | c / a+b+c+d  | | a+b / a+b+c+d  | b+c / a+b+c+d  | a / a+b+c+d  | c / a+b+c+d  |
  
  
-x1sq - x1psq  ~= x2sq - x2psq+x1zsq - x1spsq  ~= x2zsq - x2spsq
 0.311211 ~= 0.311583 0.311211 ~= 0.311583
 +
 +아래는 r 에서 계산한 것
 +<code>
 +> .794^2 - .565^2
 +[1] 0.3112
 +> .692^2 - .409^2
 +[1] 0.3116
 +</code>
  
 R에서 보는 예는 아래를 참조 R에서 보는 예는 아래를 참조
Line 269: Line 279:
  
 <code> <code>
-pcor.test(datavar$bankaccount, datavar$income, datavar$famnum) +pp.b.i <- pcor.test(datavar$bankaccount, datavar$income, datavar$famnum) 
-pcor.test(datavar$bankaccount, datavar$famnum, datavar$income)+p.b.i 
 +p.b.i$estimate 
 + 
 +p.b.f <- pcor.test(datavar$bankaccount, datavar$famnum, datavar$income) 
 +p.b.f 
 +p.b.f$estimate 
 + 
 +sp.b.i <- spcor.test(datavar$bankaccount, datavar$income, datavar$famnum) 
 +sp.b.i 
 +sp.b.i$estimate 
 +sp.b.f <- spcor.test(datavar$bankaccount, datavar$famnum, datavar$income) 
 +sp.b.f 
 +sp.b.f$estimate 
 + 
 + 
 +zc.b.i <- cor(datavar$bankaccount, datavar$income) 
 +zc.b.i  
 +zc.b.f <- cor(datavar$bankaccount, datavar$famnum) 
 +zc.b.f 
 + 
 +zc.b.i^2 - (sp.b.i$estimate)^2 
 +zc.b.f^2 - (sp.b.f$estimate)^2
  
-spcor.test(datavar$bankaccount, datavar$income, datavar$famnum) 
-spcor.test(datavar$bankaccount, datavar$famnum, datavar$income) 
 </code> </code>
 . . .  . . . 
 <code> <code>
-> pcor.test(datavar$bankaccount, datavar$income, datavar$famnum) +> pp.b.i <- pcor.test(datavar$bankaccount, datavar$income, datavar$famnum) 
-   estimate    p.value statistic  n gp  Method +> p.b.i 
-1 0.7825112 0.01267595  3.325102 10  1 pearson +  estimate p.value statistic  n gp  Method 
-> pcor.test(datavar$bankaccount, datavar$famnum, datavar$income) +1   0.7825 0.01268     3.325 10  1 pearson 
-   estimate    p.value statistic  n gp  Method +> p.b.i$estimate 
-1 -0.672856 0.04702022 -2.406425 10  1 pearson +[1] 0.7825 
-> +>  
-> spcor.test(datavar$bankaccount, datavar$income, datavar$famnum) +> p.b.f <- pcor.test(datavar$bankaccount, datavar$famnum, datavar$income) 
-   estimate  p.value statistic  n gp  Method +> p.b.f 
-1 0.5646726 0.113182  1.810198 10  1 pearson +  estimate p.value statistic  n gp  Method 
-> spcor.test(datavar$bankaccount, datavar$famnum, datavar$income) +1  -0.6729 0.04702    -2.406 10  1 pearson 
-    estimate   p.value statistic  n gp  Method +> p.b.f$estimate 
-1 -0.4086619 0.2748117 -1.184655 10  1 pearson+[1] -0.6729 
 +>  
 +> sp.b.i <- spcor.test(datavar$bankaccount, datavar$income, datavar$famnum) 
 +> sp.b.i 
 +  estimate p.value statistic  n gp  Method 
 +1   0.5647  0.1132      1.81 10  1 pearson 
 +> sp.b.i$estimate 
 +[1] 0.5647 
 +> sp.b.f <- spcor.test(datavar$bankaccount, datavar$famnum, datavar$income) 
 +> sp.b.f 
 +  estimate p.value statistic  n gp  Method 
 +1  -0.4087  0.2748    -1.185 10  1 pearson 
 +> sp.b.f$estimate 
 +[1] -0.4087 
 +>  
 +>  
 +> zc.b.i <- cor(datavar$bankaccount, datavar$income) 
 +> zc.b.i  
 +[1] 0.7944 
 +> zc.b.f <- cor(datavar$bankaccount, datavar$famnum) 
 +> zc.b.f 
 +[1] -0.6923 
 +>  
 +> zc.b.i^2 - (sp.b.i$estimate)^2 
 +[1] 0.3123 
 +> zc.b.f^2 - (sp.b.f$estimate)^2 
 +[1] 0.3123 
 +> 
 >  > 
 >  > 
-  
 </code> </code>
  
Line 586: Line 641:
 ====== e.g. 5: Stock Market ====== ====== e.g. 5: Stock Market ======
 see [[:r:multiple_regression#partial_semi-partial_correlation_and_r_squared_value|Partial and semipartial example in r]] see [[:r:multiple_regression#partial_semi-partial_correlation_and_r_squared_value|Partial and semipartial example in r]]
 +
 +====== e.g. 6: SWISS ======
  
  
sequential_regression.1606802664.txt.gz · Last modified: 2020/12/01 15:04 by hkimscil
Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki