Table 1.
Summary of main motives to gossip (from literature).
Fig 1.
Our framework of social information processing.
The central agent goes through a series of internal processes (black rectangles) after learning about some event involving a target individual, which includes accessing and updating relevant knowledge (blue rounded rectangles), and determining whether to take action about it. See text and Lee, Kralik, Jeong [72–75] for details. (Three face images from: Freepik.com).
Fig 2.
Social domains used in the current study.
Eight different social domains (with their positive and negative components) used in the current study and their relative positions from lower to higher significance, in terms of right and wrong, good and bad behavior, and potential impact. Five morality domains (prosociality, fairness, community, respect, and purity) were adopted and modified from the moral foundations theory [76–79]. The remaining three (competition, social-oriented, and general social affairs) were selected to represent other social activities.
Table 2.
Three main components differentiate the target groups.
Ingroup members (i.e., close friends) are expected to be high in intimacy and interest, but relatively low in influence; outgroup members (i.e., strangers) low in all categories; and celebrities high in influence, midlevel in interest, and low in intimacy.
Table 3.
Two main social goals as problems to solve.
Key benefits and costs for achieving the goals are listed.
Table 4.
Gossip categorization and examples.
Contents with superscript “m” contain moral aspects based on the moral foundations theory [76–79].
Fig 3.
The experimental task involved a total of seven screen displays on a computer monitor: (1) a gossip scenario was shown for seven seconds; (2) participants were then asked to decide whether or not to spread the given information to other people by pressing buttons on a keyboard; (3) if they chose to spread the gossip, they were given four options of gossip receivers: family members or a few close friends, multiple acquaintances, anyone including a total stranger, and strangers only, with receiver type to be examined in a future article; and (4)-(7) the participants were then asked to rate each scenario (from -3 to +3) on emotion (4), ordinariness (5), interest level (6), and subjective valence (7).
Fig 4.
Mean spread rates of target, valence, and content of gossip (i.e., main effects).
(A) Mean spread rates of ingroup, outgroup, and celebrity gossip. (B) Mean spread rates of positive and negative gossip. (C) Mean spread rates of eight different contents of gossip (see S1 Table in S1 Text for statistical significance): PRSC (prosociality), FAIR (fairness), PURI (purity), COMM (community), RESP (respect), CMPT (competition), SOCL (social-oriented), GSA (general social affairs). ** p < 0.01, *** p < 0.001.
Fig 5.
Mean gossip spread rates of 2-way interaction ‘target x valence’.
(A) Mean spread rates of positive and negative gossip of ingroup target. (B) Mean spread rates of positive and negative gossip of outgroup target. (C) Mean spread rates of positive and negative gossip of celebrity target. (D) Mean spread rates of positive ingroup, outgroup, and celebrity gossip. (E) Mean spread rates of negative ingroup, outgroup, and celebrity gossip. ** p < 0.01, *** p < 0.001.
Fig 6.
Mean gossip spread rates of 2-way interaction ‘content x valence’.
(A)-(H) Mean spread rates of positive and negative gossip for eight different content domains: prosociality (care/harm), fairness (fair/cheating), purity (sanctity/degradation), community (group loyalty/betrayal), respect (authority/subversion), competition (positive/negative), social-oriented (altruism/selfishness), and general social affairs (positive/negative). White bars represent positive and black bars represent negative counterparts of each content domain. * p < 0.05, *** p < 0.001.
Fig 7.
Mean gossip spread rates of 2-way interaction ‘content x target’.
(A)-(C) Mean spread rates of eight different content domains for ingroup, Outgroup, and celebrity target, respectively. (PRSC: prosociality, FAIR: fairness, PURI: purity, COMM: community, RESP: respect, CMPT: competition, SOCL: social-oriented, GSA: general social affairs.) See S2 Table in S1 Text for statistical significance.
Fig 8.
Mean gossip spread rates of 2-way interaction ‘content x target’.
(A)-(C) Mean spread rate of ingroup, outgroup, and celebrity gossip for social-oriented, general social affairs, and purity, respectively. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 9.
Mean gossip spread rates of 3-way interaction ‘content x valence x target’.
(A)-(C) Mean spread rates of content domains for positive gossip of ingroup, outgroup, and celebrity targets, respectively. (D)-(F) Mean spread rates of content domains for negative gossip of ingroup, outgroup, and celebrity targets, respectively. (CARE: care, FAIR: fairness, SANC: sanctity, LOYA: group loyalty, AUTH: authority, CMPT+: positive competition, ALTR: altruism, SOCL+: positive general social affairs for positive, and HARM: harm, CHEA: cheating, DEGR: degradation, BETR: betrayal, SUBV: subversion, CMPT-: negative competition, SELF: selfishness, SOCL-: negative general social affairs for negative counterparts of 8 contents.) See S3 Table in S1 Text for statistical significance.
Fig 10.
Mean gossip spread rates about ingroup targets for positive versus negative events.
The graphs with green shade ((A), (B), (F) & (G)) are those with higher positive spreading rates than negative spreading rates. The graphs with red shade ((C) & (D)) are those with the opposite pattern. The graphs with pale green (E) and pale red (H) are not statistically significant yet show the trends of the graphs with the corresponding colors. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 11.
Mean gossip spread rates about outgroup targets for positive versus negative events.
The graphs with green shade ((B) & (G)) are those with higher positive spreading rates than negative spreading rates. The graphs with red shade ((C) & (E)) are those with the opposite pattern. The graphs with pale green (A) and pale red ((D), (F) & (H)) are not statistically significant yet show the trends of the graphs with the corresponding colors. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 12.
Mean gossip spread rates about celebrity targets for positive versus negative events.
The graphs with red shade ((B), (C), (E) & (F)) are those with higher positive spreading rates than negative spreading rates. The graph with pale red (H) is not statistically significant yet has the trend of the graphs with its corresponding color. The graphs with pale green ((A) & (G)) are not statistically significant yet show the opposite trend of the graphs with red shade. * p < 0.05, ** p < 0.01, *** p < 0.001.