Skip to main content

Table 6 Contribution of Individual, Specialty, and Unit-Level Variables using R 2

From: Assessment of variation in the alberta context tool: the contribution of unit level contextual factors and specialty in Canadian pediatric acute care settings

Dependent Variable

Explained variance (R 2)

 

Null vs. Model 11

Model 1 vs. Model 22

Model 1 vs. Model 33

Model 2 vs. Model 34

Leadership

7.16%

14.02%

33.89%

23.11%

Culture

35.81%

6.85%

98.05%

97.90%

Evaluation

1.11%

16.62%

77.70%

73.25%

Social Capital

14.37%

-7.57%

87.62%

88.49%

OS-Staffing

20.08%

1.86%

16.85%

15.27%

OS-Space

2.72%

-1.30%

46.87%

47.55%

OS-Time

18.07%

1.19%

44.22%

43.55%

Formal interactions

20.52%

16.73%

29.98%

15.91%

Informal

interactions

24.46%

-4.91%

67.44%

68.96%

Structural and

electronic resources

91.69%

-7.71%

41.73%

45.90%

  1. 1 R 2 = 1 - (Ï„1/Ï„0) where Ï„0 and Ï„1 is the estimated unit-level error variance for the null model and model 1. This measure implies the contribution of individual level covariates in terms of relative error variance reduction when individual level covariates were added to null model.
  2. 2 R 2 = 1 - (Ï„2/Ï„1) where Ï„1 and Ï„2 is the estimated unit-level error variance for models 1 and 2. This measure implies the contribution of specialty in terms of relative error variance reduction when specialty was added to model 1.
  3. 3 R 2 = 1 - (Ï„3/Ï„1) where Ï„1 and Ï„3 is the estimated unit-level error variance for the model 1 and 3. This measure implies the contribution of specialty and unit level relative covariates in terms of error variance reduction when specialty and unit level covariates were added to model 1.
  4. 4 R 2 = 1 - (Ï„3/Ï„2) where Ï„2 and Ï„3 is the estimated unit-level error variance for the model 2 and 3. This measure implies the contribution of specialty and unit-level relative covariates in terms of error variance reduction when unit-level covariates were added to model 2.
  5. Note: negative R 2 reported in-text as '0'