Page 150 - GHES-2-3

Global Health Economics and
Sustainability
Parenting in a changing climate

consisting of experience of climate change (Items 14 – whether the DVs differed across each IV. We checked for
16, Cronbach’s alpha = 0.82), and behavioral engagement reliability with the McDonald’s Coefficient Omega (0.852,
(Items 17 – 22, Cronbach’s alpha = 0.81). The demographic mean [M] = 48.08, standard deviation = 9.632, coefficient
variables were designed using recommendations from the of variation = 92.768). An effect size ( = ÷r z N ) was
Government Statistical Service (Civil Service, 2024). All calculated for the Mann–Whitney U tests, and ()=(−
2
variables are presented in Table 1. +1)/(−) for Kruskal–Wallis and determined as either
small, medium, or large. The results are summarized in
2.2. Data collection
Table 2.
Snowballing techniques were adopted as a convenient,
low-cost, and efficient way to access parents with at least 3. Results
one child aged 5 – 11 years on social media platforms 3.1. Demographic variables
(Leighton et al., 2021). Informed consent was obtained
before participants responded to the questionnaire. Ethical There was a significant difference in the category “cognitive
approval was obtained by the Institution’s College Ethics and emotional impairment” across different participant
Committee for full consent and a questionnaire. There ages (H[2] = 6.556; p = 0.038) with a small effect size
were 153 out of 196 participants who completed the full (0.03). Younger parents (<35 years) had a higher mean
questionnaire and were included in the analysis. rank (MR) = 85.73 compared to their older counterparts,
with those aged 36 – 45 years having an MR of 66.16 and
2.3. Analysis those >45 years having an MR of 76.44. A significant

The data were analyzed using IBM Statistical Package for difference was also found in “climate change anxiety”
the Social Sciences 28 software to examine relationships between participants with a qualification of “degree level or
across and within the pre-determined characteristics. We higher” (median [Md] = 2.21; n = 89) and those with other
used null hypotheses to test for significant differences qualifications (Md = 2.07; n = 32). The Mann–Whitney
between variables. The normality test by Kolmogorov– U test results were U = 1054.5; z = −2.173; p = 0.03, with
Smirnov indicated the scale was not evenly distributed a small effect size (0.2). Similarly, there was a significant
(0.080, p = 0.018). Therefore, non-parametric statistical difference in the category “experience of climate change”
tests were performed (Mann–Whitney U test for two- based on educational qualifications. Participants with
group comparisons or Kruskal–Wallis for more than two- a “degree level of higher” had Md = 2; n = 88 compared
group comparisons). Dependent variables (DVs) were to those with “other” qualifications (Md = 1.67; n = 32;
climate change anxiety (cognitive-emotional impairment, U = 1022; z = −2.329; p = 0.02), with a small effect size
functional impairment) and eco-behaviors (experience of (0.21). There was a significant difference in the category
climate change, behavioral engagement), and independent “experience of climate change and geographical area,”
variables (IVs) were demographic variables. We analyzed based on geographical area (H[3] = 13.075; p = 0.0004)
with a small/moderate effect size (0.06). Participants living
Table 1. Categories of key variables in an “inner-city” area ranked higher (MR = 98.18) than
those in their more rural counterparts (MR: village = 82.51;
Categories Variables Instruments suburbs = 78.32; town = 60.22).
Climate (i) Cognitive-emotional Items 1 – 13 in the Climate
change impairment Change Anxiety Scale 3.2. Child/children talks about climate change
anxiety (ii) Functional impairment There was a significant difference in “climate change
Eco-behavior (i) Experience of Items 14 – 22 in the anxiety” between participants who reported “yes” that
climate change Climate Change Anxiety
(ii) Behavioral engagement Scale their child/children talks about climate change (Md = 2.3;
Demographic (i) Age Questionnaire informed n = 78) and those who reported “no” that their child/
children does not (Md = 2.04; n = 75; U = 1870.5; z = −3.85;
variables (ii) Relationship to child by the Government
(iii) No. of children Statistical Service p < 0.001), with a medium effect size (0.31). There was also
(iv) School type a significant difference in “functional impairment” (Md:
(v) Ethnic group “yes” = 1.4; “no” = 1; U = 2172; z = −2.754; p = 0.006) with
(vi) Education a small effect size (0.23), “experience of climate change”
(vii) Employment status
(viii) Geographical (Md: “yes” = 2.33; “no” = 1.67; U = 2136.5; z = −2.812;
location p = 0.005) with a small effect size (0.22), and “behavioral
(ix) Child’s climate change engagement” (Md: “yes” = 4.17; “no” = 3.83; U = 1920;
(x) Education z = −3.580; p < 0.001) with a small effect size (0.29). This

Volume 2 Issue 3 (2024) 3 https://doi.org/10.36922/ghes.3172