*2.5. Data Analysis*

Descriptive and inferential statistical analyses were subsequently performed. Chi-square (χ2) was applied to compare the distribution of responses to the discrete outcome variable among two or more independent comparison groups. The χ2 test provides information not only on the significance of any observed differences, but also provides detailed information on exactly which categories account for any differences found. The χ2 test was performed to compare demographic characteristics between vendor and non-vendor groups, as well as the difference in percentages of vendors/non-vendors' and male/females' responses, in terms of health problems associated with excessive lighting. Fisher's exact test was instead used in cases where there were cells with a frequency less than 5 and/or with small group sizes, as the Fisher's exact test has no sample-size restriction.

Two sample T-tests were performed with continuous data and Likert data to compare the difference in means between the two independent groups' responses, i.e., between vendors and non-vendors, and between male and female, regarding respondents' demographic characteristics, awareness of excessive light impacts on human health, and attitude toward the contribution of night markets to local development. Similarly, the one-way analysis of variance (ANOVA) was used to explore the differences in means among more than two groups (i.e., age groups).

Some statistical notations used in the paper are reported in Table 1.


**Table 1.** Statistical notations.

### **3. Results and Discussion**
