Skip to main content

Table 3 Research use domains, subactions, subaction levels, raw utilities, standard errors, and rescaled utility coefficients for each domain

From: Using conjoint analysis to develop a system of scoring policymakers’ use of research in policy and program development

Research use domain

Key subaction

Levels of each subaction

Raw utility coefficient (SE)a

Rescaled utility coefficientb

Relative importance (%)

1. Conceptual research use

a. Background understanding of the health issue

i. No

−0.92 (0.15)

0

11.47

ii. Yes

0

1.19

b. Understanding of the policy context

i. No

−1.09 (0.14)

0

16.27

ii. Yes

0

1.41

c. Value, skills, and/or understanding of research use in policy

i. No

−0.84 (0.16)

0

9.82

ii. Yes

0

1.09

d. Alternative perspectives and/or strategies

i. No

−1.26 (0.14)

0

18.94

ii. Yes

0

1.62

e. Informed core understanding of the issue

i. No, just clarified

−0.69 (0.12)

0

9.43

ii. Yes

0

0.88

f. Examples were specified

i. No

−2.18 (0.23)

0

34.07

ii. Yes

0

2.81

2. Instrumental research use

a. Research informed the core of the decision

No/negligible influence

−4.34 (0.50)

0

69.37

ii. No, additional details only

−2.37 (0.33)

2.15

iii. Yes

0

4.75

b. Examples were specified

i. No

−1.92 (0.27)

0

30.63

ii. Yes

0

2.10

3. Tactical research use

a. Support, confirm, or justify predetermined decisions

i. No

−1.44 (0.19)

0

24.59

ii. Yes

0

2.21

b. Persuade targeted stakeholders

i. No

−1.88 (0.22)

0

32.25

ii. Yes

0

2.90

c. Persuade peripheral stakeholders

i. No

−1.10 (0.15)

0

18.89

ii. Yes

0

1.70

d. Inform stakeholders about key issues

i. No

−1.42 (0.18)

0

24.26

ii. Yes

0

2.18

4. Imposed research use

a. Organisation mandates research use

i. No

−3.37 (0.60)

0

43.69

ii. Yes

0

3.93

b. Organisation expects research use

i. No

−2.64 (0.44)

0

34.31

ii. Yes

0

3.09

c. Organisation encourages research use

i. No

−1.70 (0.39)

0

22.00

ii. Yes

0

1.98

  1. aAll P <0.001 (using Wald χ2 statistics).
  2. bUtility coefficients were rescaled so that they are positive, that the lowest level of each subaction has a zero-coefficient, and add up to 9.