NEHA March 2024 Journal of Environmental Health

able linear regression modeling was used to investigate factors related to protective parental behaviors for preventing THS exposure. Poten- tial interaction eects were explored to con- sider the extent to which child age moderated THS-related knowledge, attitudes, and e- cacy-based perceptions; none was significant, indicating that the age of the children was not a main eect on THS-related protective behav- iors. Similarly, interaction eects with smoking status were also not found. Reported analyses were completed in Stata version 15.1.

Demographics, Smoking Behaviors, and Characteristics Known to Be Related to Thirdhand Smoke Exposure Among Parents in California ( N = 363) TABLE 1 continued

Age Groups of Children in the Household*

Overall # (%)

All Children <13 Years ( n = 120)

All Children 13–17 Years ( n = 136)

All Children <18 Years ( n = 107)

Employment Full-time

Results

42 28 13 17 20 46 58 16

30 25 13 36 32 49 56 31

33 22

105 (29)

Part-time

75 (21)

Participant Characteristics In our sample of 363 parents, the majority self-reported female (84%), White (66%), had at least some college education (64%), lived in a single-family home (68%), and did not own the home they were living in (63%). Almost one third (30%) of the participants identified as Hispanic or Latinx. The aver- age age of participants was 36.57 years ( SD = 16.09). Most participants did not smoke (93%), and 62% reported that they had never heard of THS. Households had, on average, two children <18 in the home ( M = 2.08, SD = 1.75). Demographic characteristics by number of children in the household are pre- sented in Table 1, with mutually exclusive households showing homes with only chil- dren <13 years, with only adolescents (i.e., 13–17 years), and with children ages 17 and younger (i.e., at least one child <13 years and at least one adolescent). To examine the role of child age on THS perceptions, we compared parents who had children of dierent ages. Findings displayed in Table 2 show that, compared with par- ents who had only young children (i.e., <13 years) living in their home, parents who had only older children (i.e., 13–17 years) had significantly lower THS-related ecacy and behaviors, with no statistical dierence in knowledge or attitudes. Compared with par- ents who had only young children living in their home, parents who had both younger and older children had not only significantly lower THS-related ecacy and behaviors but also significantly lower knowledge and atti- tudes. Thus, the first hypothesis is partially supported in that parents who had only younger children living in their homes had significantly higher THS-related knowledge,

Retired

30 (8)

Unemployed

25 23

78 (21) 75 (21)

Other

Home ownership Own

38 48 21

133 (37) 162 (44)

Rent

Other

68 (19)

Home type Apartment, condominium, townhome

96 (26)

Single-family home

247 (68)

Other

20 (6)

Traveled in last year Yes

107

116

88 19

311 (86)

52 (14)

Rented a car in last year Yes

43 77

47 89

31 75

121 (33) 241 (67)

Outcome variables in M (SD) Knowledge

4.51 (0.54) 4.23 (0.69) 3.75 (0.83) 3.98 (0.94)

4.36 (0.64) 4.09 (0.78) 3.45 (0.93) 3.73 (0.89)

4.23 (0.88) 3.93 (0.96) 3.35 (0.96) 3.71 (0.96)

4.37 (0.64) 4.04 (0.80) 3.57 (0.90) 3.79 (0.94)

Attitudes

Efficacy

Protective behaviors

* Participants were categorized as follows: all children <13 years only, all children between the 13 and 17 years only, and all children <18 years (where at least one child is between 13 and 17 years and at least one child is <13 years).

THS exposure were also collected (e.g., home ownership, type of housing, recent travel, recent car rental). Data Analysis Participants with ≥6 missing item responses across the measures of knowledge, attitude, ecacy, and behavior were excluded from the analysis ( n = 6, 0.4%). For participants with

<6 missing responses ( n = 26, 2.4%), hot deck imputation (Andridge & Little, 2010) in Stata was used, replacing missing values via vari- able matching based on gender and smoking status. This approach replaces the missing data with a randomly selected response from the matched options. Dierences between the ages of children living in a home were explored via regression modeling. Similarly, multivari-

March 2024 • Journal of Environmental Health