4.2. Procedure and Materials
After completing an online informed consent form, participants first received the manipulation of exposure to air pollution. Participants in the air pollution condition were asked to recall a recent experience on a hazy and polluted day. Then, they used at least 50 words to describe the weather and how they felt on that day. Participants in the control condition were asked to choose one day in the past week and describe what they had done on that day with at least 50 words. We checked participants’ answers and found that all the participants followed the instructions.
Participants then reported their current moods. In addition to the six items used in Study 1, two additional items were added (i.e., angry and anxious; 1 = not at all, 7 = very much). Scores on three positive-mood items (M = 4.05, SD = 1.47, α = 0.89) and five negative-mood items (M = 2.93, SD = 1.56, α = 0.93) were averaged.
Six moral violations were adapted from previous studies to measure moral judgment [20
], including, “In order to attend an important interview on time, one doesn’t help a stranger fainted on the roadside”, “In order to graduate, one plagiarizes graduation thesis from others’ papers”, “One forbids his wife/her husband to wear clothing that he/she has not first approved”, “One tries to undermine all of his/her boss’ ideas in front of others”, “One always said publicly that his/her own home country is worse than other countries”, and “One gets married to his/her cousin.” For the two vignettes of non-moral negative behaviors, we retained the vignette of laziness from Study 1 and created another vignette, “One paints his rooms’ walls blue although all of his friends disapprove of it”.
Participants reported how acceptable they thought these behaviors were (1 = definitely unacceptable
, 7 = definitely acceptable
). The acceptance of six moral violations (M
= 3.20, SD
= 0.93, α
= 0.41) and those of two non-moral negative behaviors (M
= 5.66, SD
= 1.39, r
= 0.37) were averaged (please refer to Appendix C
for the possible explaination of the low internal consistency or correlation).
Two tasks were used to measure immoral behavioral intentions. In the first task, participants read six hypothetically unethical behaviors adopted from previous research [15
], including “Use office supplies, Xerox machine, and stamps for personal purposes”, “Make personal long-distance phone calls at work”, “Waste company time surfing on the internet, playing computer games, and socializing”, “Take merchandise and/or cash home”, “Abuse the company expense accounts and falsify accounting records”, and “Receive gifts, money, and loans (bribery) from others due to one’s position and power”. Participants reported the possibility that they would engage in these behaviors (1 = not likely at all
, 7 = very likely
). The scores were averaged to indicate their immoral behavioral intentions (M
= 2.61, SD
= 1.16, α
The second task was to allocate two tasks between oneself and other people [21
]. Participants were told that they would complete a five-minute task at the end of the research, and there were two alternative tasks. One task was easier and more interesting than the other. Participants were told that the program would randomly choose some decision-makers. If chosen as a decision-maker, the participant could distribute one task to oneself and leave the other to one of the following participants; otherwise, they would complete a task left by a previous decision-maker. Participants were also told that their roles and choices were anonymous. Actually, all the participants were designated as decision-makers. If assigning the easier task to oneself, participants’ behavior would be categorized as immoral and selfish.
After completing the above measures, participants evaluated how dirty the weather was in their recalled scenarios (1 = very dirty, 7 = very clean), with lower scores indexing more perceived dirtiness (M = 3.92, SD = 2.04). The question intended to check the effectiveness of air pollution manipulation.
As in Study 1, we asked participants to evaluate the current weather (1 = very bad
, 7 = very good
= 4.68, SD
= 1.46). Given that washing behaviors might influence their moral judgment and moral behaviors [6
], they also reported whether they had washed themselves (e.g., washing hands, having a bath, etc.) right before attending the research by choosing between yes
Then, we measured participants’ trust in government with the following statements: “I feel that government acts in citizen’s best interest”, “I feel fine interacting with the government since the government generally fulfills its duties efficiently”, “I always feel confident that I can rely on government to do their part when I interact with them”, and “I am comfortable relying on the government to meet their obligations” [22
]. Participants reported their agreements on the above items (1 = completely disagree
, 7 = completely agree
). Scores were averaged, with higher scores indicating more trust in government (M
= 4.09, SD
= 1.57, α
We further used the 4-item Perceived Awareness of the Research Hypothesis scale (PARH; [23
]) to measure whether participants had detected any hypothesis of the research. Sample items included, “I knew what the researchers were investigating in this research”, and “I wasn’t sure what the researchers were trying to demonstrate in this research” (1 = completely disagree
, 7 = completely agree
). Scores were reversed, if necessary, and averaged. Higher scores indicated more detection of the research hypothesis (M
= 3.84, SD
= 1.38, α
Finally, participants reported their genders and ages. As Study 1 did not find social class as a significant confounding variable, we did not include it in Study 2. Participants received debriefing after completing the whole research.