Skip to main content

Table 3 Parameter estimation of structural equation modeling for MNP adherence

From: Adherence to micronutrient powder for home fortification of foods among infants and toddlers in rural China: a structural equation modeling approach

Rankinga

Behavioral Input

Unstd. βb

S.E.

Z

Std. βc

Total effectsd(CI95)

Cohen’s f 2 e

1

Salience

0.377**

0.026

14.416

0.440

0.440(0.372, 0.501)

0.227

2

Intention

0.342**

0.027

12.581

0.374

0.374(0.308, 0.433)

0.156

3

Knowledge and skills

0.138**

0.019

6.428

0.214

0.214(0.120, 0.264)

0.058

4

Personal agency

0.902**

0.086

1.448

0.460

0.172(0.119, 0.218)

0.033

5

Environmental constraints

-0.476**

0.096

-4.947

-0.142

-0.142(-0.194, -0.090)

0.030

6

Caregiver generation

0.282*

0.068

4.159

0.119

0.119(0.052, 0.165)

0.029

7

Infant and toddler age

0.019*

0.007

2.839

0.081

0.081(0.022, 0.143)

0.011

8

Injunctive norm

0.413**

0.074

5.547

0.195

0.073(0.043, 0.105)

0.010

9

Family economic status

-0.063*

0.028

-2.285

-0.065

-0.065(-0.146, -0.022)

0.005

10

Instrumental attitude

0.069**

0.015

4.432

0.135

0.050(0.022, 0.082)

0.003

11

Descriptive norm

0.106*

0.040

2.671

0.086

0.032(0.007, 0.060)

0.001

  1. aThe ranking is based on the total effect of each factor on MNP adherence behavior in descending order of magnitude. bUnstd.β: Unstandardized path coefficient. cStd. β: Unstandardized path coefficient. dThe total effect of the behavioral input on the MNP adherence. For indirect factors (i.e., personal agency, injunctive norm, instrumental attitude, descriptive norm), the total effect is mediated by the intention. eCohen’s f 2 < 0.02, Cohen’s f2 ≥ 0.02, Cohen’s f2 ≥ 0.15, and Cohen’s f2 ≥ 0.35 represent very small, small, medium, and large effect sizes, respectively. *P < 0.05 **P < 0.001.