Skip to main content

Table 4 CVI-representativeness and the modified kappa statistics (κ*) of item representativeness

From: A portable mnemonic to facilitate checking for cognitive errors

Quadrant and item

CVI-representativeness

pc

Modified κ*

Evaluation of κ*

Quadrant 1: T = life or limb threat (What are the life or limb threatening conditions in this patient?)

 Item 1: cognitive errors due to failure to consider life or limb threatening conditions

1.0

0.008

1.0

Excellent

Quadrant 2: W = Wrong? (What if i am wrong? what else could it be?)

 Item 2: cognitive errors due to over attachment to a particular diagnosis

1.0

0.008

1.0

Excellent

 Item 3: cognitive errors due to failure to consider alternative diagnoses

0.86

0.055

0.85

Excellent

Quadrant 3: E = evidences (Do I have sufficient evidences for or against this diagnose?)

 Item 4: cognitive errors due to inheriting someone else’s thinking

0.86

0.055

0.85

Excellent

 Item 5: cognitive errors due to erroneous estimation or perception of prevalence

1.0

0.008

1.0

Excellent

Quadrant 4: D = dispositional factors (What are the environmental & emotional (2Es) dispositions influencing my decision?)

 Item 6: cognitive errors associated with patient characteristics (‘emotive’ influence of patient)

0.86

0.055

0.85

Excellent

 Item 7: cognitive errors associated with doctor’s affect or personality (‘emotive’ influence of doctor)

1.0

0008

1.0

Excellent

 Item 8: cognitive errors caused by impact of the workplace environment (‘environmental’)

1.0

0.008

1.0

Excellent

  1. The formula for modified kappa statistic (κ*) = (CVI-Relevance − pc)/(1 − pc), where pc represents probability of a chance occurrence [18]
  2. pc is the probability of chance of occurrence. The formula for pc is: N!/[A!*(N-A)!]*0.5 N where N = the number of judges, A = the number agreeing on good relevance [18]
  3. Evaluation criteria for modified kappa (κ*): κ* = fair (0.40–0.59), κ* = good (0.60–0.74) and κ* = excellent (>0.74)
  4. CVI should be 0.88 and above to establish validity with a p < 0.05