Skip to main content

Table 4 Monte Carlo experiment for four practices and 5,000 observations (10,000 draws)

From: A note on testing for complementarity and substitutability in the case of multiple practices

True effect σε = 0.25 σε = 1 σε = 3.5
  Compl Neither Subst Compl Neither Subst Compl Neither Subst
Multiple-restrictions test
 Complements 3,143 87 0 3,140 90 0 2,849 381 0
 Neither 1 3,534 0 2 3,531 2 654 2,277 604
 Substitutes 0 82 3,153 0 86 3,149 5 358 2,872
 Correct (%)   98.30%    98.20%    79.98%  
All cross-term test
 Complements 3,230 0 0 3,210 20 0 2,957 272 1
 Neither 1,651 342 1,542 1,537 556 1,442 975 1,613 947
 Substitutes 0 0 3,235 0 28 3,207 0 289 2,946
 Correct (%)   68.07%    69.73%    75.16%  
Single cross-term test
 Complements 2,944 72 214 2,919 115 196 2,745 359 126
 Neither 1,599 433 1,503 1,486 666 1,383 961 1,639 935
 Substitutes 202 68 2,965 184 126 2,925 95 424 2,716
 Correct (%)   63.42%    65.10%    71.00%