# Table 2 Regression models to predict EBP behavior scores over time in different groups of respondents based on their information seeking patterns

Regression coefficients

Model 1

Model 2

Model 3

Respondents

Seeking information from a peer who is:

Â Â Â

Baseline

â€ƒNot highly engaged

Not highly engaged

9.1(0.7)

9.1(0.7)

10.5(0.5)

Highly engaged, but not central

âˆ’3.2(1.4)*

Highly engaged central

âˆ’0.7(1.1)

â€ƒHighly engaged, but not central

Not highly engaged

1.6(1.1)

âˆ’0.3(1.4)

1.5(2.3)

Highly engaged, but not central

âˆ’4.3(3.0)

Highly engaged central

âˆ’2.1(1.9)

â€ƒHighly engaged central

Not highly engaged

5.5(1.9)**

5.5(5.2)

Highly engaged, but not central

2.0(4.8)

Highly engaged central

4.3(2.1)*

Change from baseline

â€ƒNot highly engaged

Not highly engaged

0.07(0.4)

0.07(0.4)

âˆ’0.8(0.6)

Highly engaged, but not central

âˆ’0.3(1.3)

Highly engaged central

2.5(1.1)*

â€ƒHighly engaged, but not central

Not highly engaged

2.4(1.0)*

1.4(1.3)

âˆ’0.5(2.3)

Highly engaged, but not central

âˆ’1.8(3.2)

Highly engaged central

2.9(1.7)

â€ƒHighly engaged central

Not highly engaged

4.2(1.8)**

âˆ’0.3(4.7)

Highly engaged, but not central

4.8(7.5)

Highly engaged central

4.9(1.9)**

Random effects variance

â€ƒHealth Units

1.03(1.1)

1.0(1.0)

~0

â€ƒIndividuals

52.5(4.3)

51.7(4.3)

47(4.9)

â€ƒResidual

36.1(2.7)

36.0(2.7)

33(2.9)

1. 1: The coefficients for all levels at baseline represent the difference from the first row
2. 2: The coefficients for all levels at â€˜change from the baselineâ€™ represent the difference from their baseline
3. *:pâ€‰<â€‰0.05, **:pâ€‰<â€‰0.01