Skip to main content

Table 6 Cluster analysis results (N = 79)

From: Exploring unnecessary invasive procedures in the United States: a retrospective mixed-methods analysis of cases from 2008-2016

Clustering Variables Clusters Va n
1 (n = 21) 2 (n = 9) 3 (n = 17) 4 (n = 11) 5 (n = 21)
Traits/Motives
 Financial gain 100%(21) 66.7%(6) 100%(17) 81.8%(9) 95.2%(20) .42 73
 Personality disorder 100%(21) 22.2%(2) 11.8%(2) 45.5%(5) 38.1%(8) .67 38
 Poor problem solving 0% 100%(9) 0% 0% 0% .99 9
 Ambition 0% 0% 5.9%(1) 0% 9.5%(2) .22 3
Environmental factors
 Oversight deficits 100%(21) 77.8%(7) 82.4%(14) 0% 100%(21) .83 63
 Corrupt moral climate 0% 0% 0% 0% 100%(21) .99 21
 Financial COI 0% 0% 47.1%(8) 0% 14.3%(3) .53 11
 Ambiguous norms 0% 0% 17.6%(3) 9.1%(1) 0% .33 4
Summary of Clusters 1 2 3 4 5   
  Financial gain H M H M H   
  Suspected antisocial H M L M M   
  Poor problem-solving H   
  Ambition L L   
  Oversight deficits H M M H   
  Corrupt moral climate H   
  Financial COI M L   
  Ambiguous norms M L   
 Seriousness rating 54.4 (7.9) 51.9(8.2) 45.6(11.7) 50.9(7.9) 47.9(10.8)   
  1. aCramer’s V was used to test the association of these nominal variables. Values above .35 are particularly important in discriminating among clusters, which is reinforced by p-values of < .05 for those Vs; values < .35 indicate weak relationships that did not discriminate significantly
  2. Percentage is percent of cases within the cluster. The raw number of cases representing a variable appears in parentheses
  3. H High (84–100%), M Medium (17–83%), L Low (1–16%); − = Absent