A process evaluation of the Supermarket Healthy Eating for Life (SHELf) randomized controlled trial

Table 3 Process evaluation of the overall study at post-intervention

Survey questions	Control n = 161^a % (n)	Price reduction n = 161^a % (n)	Skill-building n = 160^a % (n)	Combined n = 160^a % (n)	P-value
Have you changed the way you buy, cook or eat food after taking part in this study? (effectiveness)					<0.001
Yes	14.7 (23)^b	32.9 (52)^c	34.5 (51)^c	37.5 (57)^c
No	77.6 (121)	59.5 (94)	52.0 (77)	54.6 (83)
Don’t know	7.7 (12)	7.6 (12)	13.5 (20)	7.9 (12)
(If you have children aged 12 years of younger) Do you think your child/children are more willing to eat fruit and vegetables as a result of you taking part in this study? (reach)					0.227
Yes	15.4 (14)	26.0 (19)	22.1 (15)	29.7 (19)
No	61.5 (56)	58.9 (43)	54.4 (37)	45.3 (29)
Don’t know	23.1 (21)	15.1 (11)	23.5 (16)	25.0 (16)
(If you have a partner) Do you think your partner is more willing to eat fruit and vegetables as a result of you taking part in this study? (reach)					0.017
Yes	18.1 (23)^d	34.7 (42)^e	30.6 (34)^e	37.1 (43)^e
No	59.1 (75)	45.5 (55)	46.9 (52)	38.8 (45)
Don’t know	22.8 (29)	19.8 (24)	22.5 (25)	24.1 (28)

RE-AIM constructs examined are listed in brackets after each question
Fisher’s exact tests and χ ² tests (for categorical data) and two-sided unpaired t-tests with unequal variances were conducted to assess differences in survey responses between treatment groups. Unadjusted multinomial logistic regression with robust standard errors was used to assess differences between groups where Fisher’s or χ ² tests revealed overall differences
^aRepresents highest possible sample size for each question. The sample size for each question differs due to missing responses
^b,cValues for variables without a common letter differ (p < 0.0001)
^d,eValues for variables without a common letter differ (p < 0.05)

ISSN: 1479-5868