Skip to main content

Table 1 Descriptive statistics for the analysis variables in the analytical sample for the case study, using the data from Victorian Adolescent Health Cohort Study (n=953)

From: Should multiple imputation be stratified by exposure group when estimating causal effects via outcome regression in observational studies?

Role

Label

Variable

Unexposed

Exposed

Missing (%)

   

n (%) or mean (SD)a

 

Exposure

X

Cannabis use, Yes

869 (91.2)

84 (8.8)

0

Outcome

Y

Adulthood mental health scoreb

-0.02 (0.97)

0.17 (1.30)

10.3

Confounder

\(C_1\)

Parents not completed high-school, Yes

322 (37.1)

35 (41.7)

0

Confounder

\(C_2\)

Parental divorce, Yes

169 (19.4)

38 (45.2)

0

Confounder

\(C_3\)

Antisocial behaviour, Yes

69 (7.9)

31 (36.9)

0

Confounder

\(C_4\)

Alcohol use, Yes

218 (31.1)

65 (87.8)

18.6

Confounder

\(C_5\)

Adolescent depression & anxiety, Yes

432 (56.5)

62 (81.6)

11.9

Auxiliary

A

Participant’s age at wave twob

-0.01 (1.00)

0.15 (0.95)

8.4

  1. aFor incomplete variables, the descriptive statistics are obtained from the records with available data on the given variable
  2. bIn standard deviation units, standardised to the combined sample