Table 2.

Results of regression analyses assessing the impact of the job aid intervention on the five primary outcome indicators for analysis and use of vaccine-preventable disease data

Outcome indicator	Mean values from Likert scale			Regression results
n = 87 observations	Intervention group	Pre-test	Post-test	Model*	Interaction coefficient †	Standard error	P-value
1. Perceptions of availability of quality VPD data	Intervention	2.75	3.28	Linear regression	0.6533	0.4722	0.1664
	Control	3.00	3.18
2. Perceived capability to perform analysis of VPD data	Intervention	3.40	3.90	Linear regression	1.1191	0.3110	0.0003
	Control	4.50	3.80
3. Likert scale question measuring motivation to carry out analysis	Intervention	3.16	3.57	Multinomial regression	-2.4354	1.2843	0.0579
	Control	4.09	3.50
4. Perceived value of using analyzed VPD data for decision-making	Intervention	4.08	4.09	Linear regression	0.2824	0.2205	0.2004
	Control	4.56	4.38
5. Likert scale question measuring perceived motivation to use surveillance data	Intervention	3.81	3.92	Multinomial regression	-1.4828	1.4084	0.2924
	Control	4.18	3.60

VPD = Vaccine preventable disease

*Regression models included: intervention group, subject, survey round, intervention group*survey round interaction term, and controlled for sex, age, years of experience and rayon. Linear and multinomial regression performed using generalized linear model in SAS (Proc Genmod), with generalized estimating equation (GEE) used to account for correlations between subjects with repeated measures.

^†Impact of intervention assessed by intervention group*survey round interaction term.