Department of Psychology

A study of the MUM effect on a function of the recipient’s relationship to the sender using the “lost email technique.”

Abstract:

This study looked at the MUM effect (i.e. message transmittance) in comparison to the relationship between the messenger and the potential recipient. The study involved 120 Georgetown University Students and was carried out via a lost email technique in which a “lost” email containing bad, good, or neutral news was sent to a potential messenger (either a close friend or acquaintance of the recipient). The messenger then had to decide whether or not to transmit the email to the recipient, a confederate. The findings of this study show that bad news is transmitted more often between close friends than acquaintances. These findings are important because they bring new dimensions to and further support the MUM effect as well as give an insight into the different relationships people make.

Introduction:

Since the 1970s, researchers have studied message transmission in different groups of people in an aim to make a generalization of how much we can rely on our fellow human beings to relay a message that may be displeasing. Overall, they have found that bad news is transmitted less frequently than good news; this concept has been called the MUM effect. The MUM effect stands for the tendency to keep Mum about Undesirable Messages. In the seminal paper, Rosen and Tesser (1970) explained that “the reluctance to transmit information is directly dependant on the inferred desirability of the message for the potential recipient” (p. 253); thereby defining the MUM effect.

The Rosen and Tesser (1970) study was conducted so that a message would have to be transmitted face to face. Although the results of this article supported the MUM effect, it proposes that the MUM effect may have to do with the fear of possibly being “infected with the emotional reaction or evaluation by the recipient” (Rosen & Tesser, 1970, p. 262). A follow up study by Tesser, Rosen, and Batchelor (1972) supported this hypothesis (and the MUM effect) through a role play situation. A field study, also supporting the MUM hypothesis, looked at message transmittance in a group of people that gave aid to the disabled found that aid givers took longer to communicate denial of aid than assurance of aid (Tesser et al., 1971). Others have tried to define this effect in different ways. In a study done by Heath (1996), the distinction between good news and bad news is dropped, and instead he puts messages in two different categories, surprising and not surprising. He shows that surprising news is transmitted at a lower frequency than non-surprising news.

Another study done by Tesser et al. (1972) looked at “message variables and the MUM effect.” It showed that the intensity of the news (good or bad) was directly proportional to how often it was transmitted. This study was done by simply asking participants to imagine a situation in which they had to transmit different types of news and then answer a questionnaire. Although this study supported the MUM effect, the imagining a situation method is not a very reliable way of assessing behavior. This is because what one may say he or she would do in a certain situation, may not be what he or she actually does.

Conlee and Tesser (1973) found that if the potential messenger felt that the recipient wanted to hear the news (i.e. the news was good), it did make a difference in how often both bad and good news were transmitted. Both good news and bad news had a relatively higher rate of transmittance if the recipient wanted to hear it. This study was also a face to face confrontation between the messenger and the potential recipient. Another study done by Tesser and Rosen (1972) showed that the decrease of bad news transmission has to do with the messenger feeling guilty that he or she is unable to share the same “bad news” fate as the recipient. They tested this by looking at message transmission when the messenger could empathize with the recipient and when he or she could not. They found that transmittance of bad news was higher when the messenger could empathize with the potential recipient. It was suggested that a messenger may be able to empathize more with a friend than someone they did not know. A study done by Weenig (2001) looked at this possibility.

Weenig et al.(2001) did a study on the transmittance of rumors. He suggested that bad news of different intensities was transmitted more often if the recipient was a friend of the messenger rather than a complete stranger. This study was done by a face to face confrontation between the messenger and the recipient. A problem with the previous study is that the rumors, in this case, were not true. For this reason, it is possible that messengers did not feel as though conveying the news would make a difference. Also, most of the studies mentioned above were done via a face to face confrontation between the messenger and the recipient. As proposed in the starting paper by Terser and Rosen (1970), the MUM effect may be predicated on a fear of being infected with the emotional distress of the recipient. Also, the authors suggest the potential messenger may fear having to assume an emotional state congruent with the bad news the email would transmit. When a face to face confrontation takes place, the messenger may have to deal with the above mentioned consequences; thereby leading to the MUM effect. When a rumor is supposed to be transmitted to a recipient, the messenger may not care too much about it since they know the rumors are lies. This may have been a confounding variable for Weenig’s study, also leading to the MUM effect. Seeing flaws in the studies that have already been done, I studied the MUM effect via email. This has never been done before. The good, bad or neutral news were sent to a potential messenger via email. The messengers were either a close friend or an acquaintance of a potential recipient. The email was sent by a teaching assistant who “seemed” to have simply sent the email to the wrong email address. The recipients were my “confederates” and kept track of which emails were forwarded to them. The experiment was done before spring break so that confounds, such as close friends seeing each other more and may have more of a chance to transmit the message, were decreased. A study done by Bryson and Hamblin was carried out in a similar way; they wrote 120 postcards with one of three different types of messages, good news, bad news and neutral news and put the already stamped postcards on people’s car windshields. They then wanted to see if the owners of the cars would mail the postcard although it was not meant for them. They compared transmittance to type of news on the postcard. Their results also supported the MUM effect. The results can be interpreted as supporting Tesser and Rosen’s predication (even though there was not any physical contact between messenger and recipient) that bad news is emotionally stressing and for that reason, people are more likely to simply ignore the postcard with the bad news instead of mailing it on.

This experiment will act as a check to see if Tesser and Rosen’s prediction (1970), that there is an emotional stress put on the messenger and this may be the reason for the MUM effect, was correct. It also tests Weenig’s hypothesis (2001) that bad news will be transmitted more often to close friends than acquaintances because Weenig used rumors that were not true; while this experiment allows the messenger to believe that the bad news is in fact true. When a message is transmitted by email, one simply needs to forward the email to the correct recipient. This does not require any emotional or physical contact with the recipient. It is expected that overall, the transmittance of bad news will be less than that of good news. But, a hypothesis of whether a message containing bad news will be transmitted more frequently to a close friend or a stranger can go either way. If Tesser and Rosen’s prediction, is correct there should be no difference between the two groups, and if Weenig is correct then the rate of transmittance of bad news should be higher between close friends than acquaintances.

Method:

Participants:

The participants were college students (age 18 - 22) on the Georgetown campus.

Since this is a 3 (email types) x 2 (relationship) and 20 subjects were used in each category, there were 120 subjects in this study. There were 12 potential recipients, my “confederates,” and each chose 10 potential messengers. Out of the ten potential messengers for each recipient, five of those potential messengers were close friends of the recipient, while the other five were acquaintances of the recipient. Sex of the messengers was not considered in this experiment but the study attempted to use equal numbers of males and females.

Apparatus:

This study used a questionnaire which acted as a manipulation check for the friendship levels of the potential recipient and the messenger. A copy has been included in Appendix 1. Three different types of emails were used in this experiment; one containing good news, one containing bad news, and one containing neutral news. Copies of these emails have been included in Appendix 2. The informed consent form that was used is included in Appendix 3.

Design:

A 3 (email type) x 2(relationship) quasi experimental design was used. The independent variable in this experiment is who receives the different types of “lost” emails. The dependent variable in this experiment is which emails are forwarded to their potential recipients. A neutral email message served as one of the controls for this experiment. The rate at which this neutral email message is transmitted served as a baseline for how often an email would be transmitted, independent of its subject material. The rate of transmittance of the bad news and the good news was compared to that of the neutral news. Another confounding variable is that the good friend may see the potential recipient more than an acquaintance. To try and control for this, the emails were sent right before students left for spring break. In this way, there was a decreased chance that students were able to interact with one another (note: it does not eliminate the possibility completely). Emails containing good, bad, or neutral news were sent to potential messengers randomly. There were 120 subjects; 20 subjects in each group. There were 12 potential recipients and each one chose 10 messengers (five were close friends and five were acquaintances). Each potential recipient filled out a manipulation check that decided if he or she knows the messenger as a good friend or an acquaintance.

Procedure:

I chose 12 college students. These students were the potential recipients and my confederates. Each of the 12 potential recipients was asked to give names of five Georgetown students they consider close friends and five Georgetown students they consider acquaintances. These students acted as the potential messengers; 120 in all. Each potential recipient was asked to fill out the manipulation check on how well they know the messengers they have chosen. The highest possible composite score was 36 and the lowest score was 0. Each of the potential messengers received one of three possible emails. The possible emails were a good news email, a bad news email, and a neutral news email. The emails were sent to messengers randomly. This was done by separating the subjects into two groups, ascribed close friends and ascribed acquaintances. Then the email type was randomly assigned to each subject. The emails were sent from a yahoo email account: GtownTeachingAssistant@yahoo.com. All the emails were sent out on the first day of spring break in March. Data collection continued for two weeks. The “confederates” (potential recipients) kept track of what emails they received from the messengers, as well as the mode of transmittance (phone, email, etc,) and the date of transmittance. Once the data collection period was over, all participants were debriefed via email. Since this study involved deception, informed consent forms were not given to the 120 potential messengers, however, the 12 confederates were all told about the different emails and were asked to sign consent forms.

Results:

If a potential recipient received an email by the end of the two weeks, this was considered transmittance; if not, then it was considered as no transmittance. Graphs of message transmittance overall and between the close friends and acquaintances can be seen in Appendix 4. A two way chi squared analysis was done on the data.

A chi square was also done with the data from each of the different email types, comparing only the effect of friendship on the message transmittance of each of the different types of email messages.

Since MUM effect studies in the past have used acquaintances or strangers when considering message transmittance, the following tables (# 6 and 7) consider differences in transmittance only among the acquaintance group to test the presence of the MUM hypothesis.

The chi square above (Tables 2, 3 and 4) analyzed the transmittance of the type of news related only to the friendship between the recipient and the messenger.

Graphs of message transmittance overall and between the close friends and acquaintances can be seen in Appendix 4.

A composite score was added up from the manipulation check (a questionnaire that is supposed to measure friendship), with the highest composite score being a 36 (best friends) and the lowest being 0 (complete stranger). The scores that were obtained for each of the potential messengers were recorded into either the close friend group or the acquaintance group, depending on whether the potential recipient had identified the messenger as an acquaintance or close friend. The means and standard errors of the two groups’ manipulation check scores were compared on the graph that has been attached in Appendix 5. The manipulation check was used to make sure that there was a significant difference between the friendship levels of acquaintance and close friend as ascribed by the potential recipient. A t-test was done on this data and a probability of 4.11308E-15 was found.

Discussion:

The results show that there is an increased transmittance of news overall among close friends than acquaintances and that the transmittance of bad news among close friends was a lot higher than among acquaintances. These results are seen in Tables 5 and 4 respectively. The results support Weenig’s theory that bad news will be transmitted more often among close friends than acquaintances. There was not a significant difference in good or neutral news message transmittance between friends and acquaintances. For this reason, it can be said that the difference in transmission of bad news between close friend and acquaintance was due to the content of the email and the difference in friendship. Although the hypothesis was supported by the results, there are limitations to these results.

Table 1, a chi square test with all the data does not show to be significant and for this reason, the other chi square tests were done to isolate the effects of friendship on the different email types. The first graph in Appendix 4, of overall transmittance, cannot be viewed as the graph for representing the MUM effect. To represent the MUM effect, the difference in email transmittance must be considered in the acquaintance group only. The reason for this is that past researchers have shown the MUM effect to occur among acquaintances or strangers. For this reason, the close friend email transmittances may obscure the data.

The results do not support the overall MUM effect, although there is a trend seen. These results can be seen in table 7. Since past MUM effect studies have been done among acquaintances or strangers, the transmittance of news among acquaintances is used to check if the data supports the MUM hypothesis overall. A trend is scene, with p < .10, that bad news is transmitted less often than good news. The reason that the MUM hypothesis was not strongly supported may be due to the small sample size and using a larger sample size would be more likely to support the MUM effect. In an effort to show the significance of the MUM effect, past studies have compared the transmittance of good news and bad news. For this reason, Table 7, which compares transmittance of good news vs. bad news among acquaintances, is an assessment of the MUM effect as done in past studies. In Table 6, which compares transmittance of neutral news to bad news, a trend is not seen. For this reason, more researchers need to compare the transmittance of neutral news to bad news because this will be a true measure of the possibility that bad news is transmitted less frequently because of its content. The results, as presented in past studies, leave a few questions for the reader to answer. If comparing only good news and bad news transmittance, a decrease in bad news transmittance could mean:

a. good news is transmitted more often, because of its content and the bad news does not necessarily have anything to do with the decreased rate of transmittance, or
b. bad news is transmitted less often because of its content.

For this reason, the MUM effect is left unclear because the reason behind it is obscure. By comparing the rate of bad news transmittance to neutral news transmittance, one can find out if the decreased rate of transmittance is actually due to the content of the message. Or, it is possible that the there is no difference in the transmittance of bad news and neutral news (baseline transmittance) and that there is only an increased transmittance of good news. This may mean that the MUM effect, bad news is transmitted less frequently than good news, may be a result of incorrectly interpreting results and might need to be modified to say that good news is transmitted more frequently in general./p>
Since email has become a primary method of communication among college students, it is inevitable that email boxes become plagued with ‘junk” mail. There is always the possibility that the potential messenger confuses the email sent to them and treats it as junk mail. This may cause the messenger to delete the email without even looking at it. It is also possible that the potential messenger did not find the email believable.

The theoretical implications of this study are that it provides further support for the MUM hypothesis, that bad news is transmitted less often than good news. The reason for this may be that receiving an email with bad news is emotionally distressing for a person. Once they have received the email, they could either ignore it or forward it to the true recipient. Ignoring the email would decrease the amount of emotional stress upon the person who received the missent email, because they know that the real recipient will find out eventually. Forwarding the email is more emotionally distressing because it causes the person forwarding the email to keep having to think about the real recipient and how it has or will affect the real recipient. Forwarding the email also means that the missent email recipient knows something about the real recipient that he should not know. For these reasons, forwarding the bad news email is emotionally distressing. The interaction is that it seems that close friends are more likely to become part of an emotionally distressing situation for each other than acquaintances. This finding supports, and in fact, expands Tesser and Rosen’s theory (1970) that bad news is transmitted less often than good news due to an emotional stress put on the potential messenger. This study expands the Tesser and Rosen (1970) hypothesis by showing that physical, face to face contact is not necessary to support the their theory.

In conclusion, this study supports both Weenig’s hypothesis, that bad news will be transmitted more often between friends than acquaintances, and Tesser and Rosen’s hypothesis that there is emotional stress placed on the messenger and for this reason there may be a decrease in the transmittance of bad news. In the future, it would be interesting to see if there is a connection between the transmittance of bad news and gender.

---

Appendicies
Appendix 1

Appendix 2

Good News

Dear ________
Hi. This is your TA from ______ class. Your professor wanted me to tell you about your performance on your most recent paper and said I could contact you at this email address:
_________________ You received an A because it was very well written and you expressed your ideas clearly and coherently! I wanted to congratulate you on your tremendous achievement. Since this paper is worth 50% of your total grade, you are doing extremely well in the class! I will be away on vacation till the Wednesday after spring break (March 19th) but I will talk to you about this in class when we get back!
~Jon

Neutral News

Dear ________
Hi. This is your TA from _______ class. Your professor told me I could contact you at this email address: _________________ . I was just informing each student personally that I am available for help and you can make an appointment with me by signing up on your prof.’s office door. Since the next paper isn’t due until mid- April, you have plenty of time to make your appointment. I will be away on spring break till the Wednesday after spring break (March 19th) but I can talk to you about this in class when we get back. Thanks!
~Jon

Bad News

Dear ___________
Hi. This is your TA from ______ class. Your professor wanted me to tell you about your performance on your most recent paper and said I could contact you at this email address:
_____________________ You received a D because it was a very poorly written paper and you did not express your ideas (and lack thereof) clearly and coherently. I’m informing you during spring break because this grade puts your class average in great jeopardy since it is worth 50% of you total class grade. I will be away on vacation till the Wednesday after spring break (March 19th) but I will talk to you about this in class when we get back.
~Jon

Appendix 3: Informed Consent Form

I, ___________________________, understand that I am participating in a psychology experiment for a Research Methods and Statistics Class. This experiment involves me receiving emails that may contain good, bad, or neutral news. I have been informed of the contents of each email so that I am not surprised when I receive these emails.
   I understand that my participation in this study is strictly voluntary and I can withdraw at any time.
   I know that the results of this study will be used for research purposes only and will remain confidential.
   I know that if I have any questions or problems regarding the study, I can contact Tanvi Parikh at 202-784-8606. If I have any additional questions regarding my rights I have been advised that I contact the professor of the course, Dr. Rachel Barr at 202-687- 6988.
   I confirm that I have received a copy of this form.


Signed _______________________

Date __________________________


Appendix 4: unavailable

Appendix 5: unavailable

 

---

References:
Conlee, M., & Tesser, A. (1973). The effects of recipient desire to hear on news transmission. Sociometrv. 36, 588 — 599.

Heath, C. (1996). Do people prefer to pass along good news or bad news? Valence and relevance of news as predictors of transmission propensity. Organizational Behavior and Human Decision Processes. 68, 79 — 94.

Rosen, S & Tesser, A. (1970). On reluctance to communicate undesirable information: The MUM effect. Sociometry. 33, 253 — 263.

Solso, R. L., & MacLin, K. M. (2002). Experimental psychology. Boston, MA: Allyn & Baycon.

Tesser, A. & Rosen, 5. (1972). Similarity of objective fate as a determinant of the reluctance to transmit unpleasant information: the MUM effect. Journal of Personality and Social Psychology, 23, 46 — 53.

Tesser, A., Rosen, S., & Batchelor, T. (1972). On the reluctance to communicate bad news (the MUM effect): A role play extension. Journal of Personality. 40, 88 — 103.

Tesser, A., Rosen, S., & Batchelor, T. (1972). Some message variables and the MUM effect. Journal of Communication, 22, 239 — 256.

Tesser, A., Rosen, S., & Tesser, M. (1971). On the reluctance to communicate undesirable messages (the MUM effect) : A field study. Psychological Reports. 29, 651 — 654.

Weenig, M., Groenboom, A., & Wilke, H. (2001). Bad news transmission as a function of the definitiveness of consequences and the relationship between communicator and recipient. Journal of Personality and Social Psychology. 80, 449 —461.