¶ … individuals are much less apt to report unpleasant messages than pleasant ones. This so-called MUM effect could be caused by self-concern, care for the other person, or worry about social norms (Rosen & Tesser, 1970). However, research on rumors indicates that people do not always seem to have such constraints and transmit bad news readily. In "Bad News Transmission as a Function of the Definitiveness of Consequences and the Relationship Between Communicator and Recipient," researchers Weenig, Groenenboom and Wilke (2001) argued it might be productive to analyze some differences between these two lines of research. They thus wanted to determine if certain factors existed that encouraged bad news transmissions. Based on Rose and Tesser's results, they decided to research through directional hypotheses if it is easier to use rumors, or second-hand communication, to transmit bad messages, as well as if someone is more apt to relate a message to one person vs. another. The researchers' first hypothesis (Hypothesis 1) was that:

news with indefinite bad consequences for the recipient would be more likely to be transmitted to the target person than news with definite bad consequences. This was based on the notion that by transmitting news with indefinite bad consequences, the communicator warns the recipient against the potential negative consequences and thus helps the recipient to prevent those potential negative consequences. In contrast, by transmitting bad news about events with definite negative consequences, as investigated by Tesser and Rosen (1975), it is much more ambiguous whether it helps the recipient.

Studies concerning factors of helping behavior have found that the more definitely a person requires assistance, and the more immediately, the more apt it is that a potential person will help. (Latane & Darley, 1970) To really assist the individual, the potential helper has to believe he/she is personally responsible to do so. This depends on such things as the availability of individuals who are in an equal or better position to offer assistance (Latane & Darley, 1968). Weenig et. al. thus assumed that the helper would see the sending of bad news with unspecified results as more helpful and crucial for the recipient than the transmission of bad news with specific consequences: That is, the earlier a person hears news with indefinite negative consequences, the more time he/she has to prevent the effects. However, with bad news and definite consequences, the time of hearing the news is not as important, since nothing can be done to prevent the negative consequences. The assumption was also made that the definitiveness of negative consequences would affect the personal costs of bad news transmission. As Tesser and Rosen had shown (1975), when people think about relating bad news about definite negative events, they anticipate little personal gain and significant costs, such as guilt or strong reactions from the person receiving the information.

Overall, the authors for Hypothesis 1 "expected that the transmission of bad news with indefinite negative consequences would elicit stronger feelings of personal moral responsibility than the transmission of news with definite negative consequences, because it would be perceived as more helpful and more urgent for the recipient, and probably also as less costly for oneself."

Weenig et. al's second hypothesis (Hypothesis 2) was that: "the likelihood of bad news transmission would be higher if the recipient is a friend of the transmitter than if the recipient is a less acquainted person." This was based on the belief that friends may define their relationship with the recipient as more suitable for the transmission of bad news than strangers and thus may feel more morally obliged to send bad news, as fewer persons would be in a better position to provide help. Lastly, their third hypothesis (Hypothesis 3) was "we expected that bad news would be transmitted less often than good news."

The hypotheses were tested in three quantitative experiments, a behavioral study and two scenario studies. In the behavioral study (Experiment 1), Hypotheses 1 and 2 were tested to find the impact of definitiveness and relationship on the transmission likelihood of bad news. Experiment 2 tested Hypotheses 1 and 2 for very bad news investigated the mediational effect of perceived moral responsibility, urgency, importance and helpfulness, and anticipated costs and benefits of transmission. The scenario described a situation where the participant would be able to transmit very bad news to a target person without being formally responsible to do so. In Experiment 3, they used the same scenario to test Hypothesis 3, concerning the MUM effect, and its potential moderators and mediators.

stranger) and definitiveness of the message's consequences (indefinite vs. definite) as independent variables. All of the news was bad. The dependent variable was whether the bad news would be transmitted." Auxiliary data and manipulation tests were also conducted for reliability/validity purposes. All analyses of covariance, ANCOVAs, reported in the article were checked for violations of assumptions concerning homogeneity of variance and interactions between covariates and experimental factors. The few violations were reported in notes. Cronbach alpha was used to determine how well the variables measured the latent construct.
Eighty pair of undergraduates from Leiden University participated in Experiment 1, with ages varying between 18 and 30. One student per dyad was asked to participate in a study on consumer decisions and to bring a close friend of the same gender. One half the dyads were kept together, the other 40 were separated to form new dyads of strangers. Within each dyad, one person was the potential communicator, or focal participant, whereas the other participant served as recipient.

The participants were made to "coincidentally" overhear bad news about another person, who was identified. The potential senders of the message were then left alone and given the opportunity to transmit the message to the recipient. Later, both parties were given a questionnaire to check the definitiveness of the message's consequences by asking:

Do you think that the person the message was intended for could do anything to influence his situation?" with 1 (yes), 2 (don't know), and 3 (no) as response categories. The relationship with the target person was checked with the question: "Did you know the person whom the message was intended for, and if yes, how would you describe your relationship with this person?," with response categories ranging from 0 (I didn't know the person) to 4 (a good friend). The perceived content of the news was measured by an open question "What did you imagine was the content of the news?" We also measured the perceived importance and urgency of the news: "How important [urgent] do you think the news was for the other person?," with response scales respectively ranging from -2 (very unimportant) to +2 (very important), and 1 (not urgent) to 4 (very urgent).

Support for Hypotheses 1 and 2 was found in Experiment 1: "Bad news with indefinite consequences was transmitted more often than bad news with definite consequences, and both kinds of bad news were transmitted more frequently when the recipient was a friend rather than a stranger." However, the news was never regarded as very urgent or important. Also, the difference between the two definitiveness conditions on perceived definitiveness news was rather small. It seemed the mildly bad news was not perceived as definite, important, and urgent enough to evoke strong reactions and analyses may have suffered because most of the variables were measured with only two or three response categories, that is, restriction of variance. Thus, the same two hypotheses were tested in a scenario study that allowed investigation worse news.

In Experiment 2, a work scenario was devised where one person received information on another person's possible dismissal from the company. Factors tested were: closeness of individuals; likelihood of transmitting information; degree of reluctance to transmit bad news. A total of 96 undergraduates participated, the age varying between 18 and 60, with a mean of 21. In Experiment 2, they found additional support for Hypotheses 1 and 2. However, because the research designs so far did not include good news, it was unknown whether the definitiveness of the news' consequences and the relationship strength also affect good news transmission. Therefore, they tested Hypotheses 1 and 2 in a second scenario study identical to that of Experiment 2, but with the added factor of the valence of the news (positive or negative). It was expected (Hypothesis 3) that bad news would be transmitted less often than good news, but no hypotheses were formulated on a possible differential effect of definitiveness and relationship on good/bad news transmission. The work scenario was repeated with bad or good news.

Hypothesis 1 was confirmed in all three experiments. In Experiment 1, the bad news with the indefinite consequences was transmitted 2. 5 times as often than bad news with definite consequences. Experiments 2 and 3 provided support to the hypothesis. Hypothesis 2 was also confirmed in all three experiments: News was transmitted almost three times as often if the recipient was a friend than a stranger. Hypothesis 3 and the…