Effect of media presentations on willingness to commit to organ donation

Significance

We often encounter cases of organ donation in the press or on television. How might these stories affect readers? We found that reading coverage of cases that include identifying information about the receiver (saved by organ donations) increased participants’ willingness to commit to organ donation themselves, to donate the organs of a deceased relative, or to support a transition to an “opt-out” policy. Conversely, identifying the deceased donor induced thoughts of death rather than about saving lives, resulting in fewer participants willing to donate organs or to support policies that facilitate organ donation. We show that most of the stories that appear in the press include an identified donor rather than an identified receiver, possibly reducing organ donations.

Abstract

We examine how presentations of organ donation cases in the media may affect people's willingness to sign organ donation commitment cards, donate the organs of a deceased relative, support the transition to an “opt-out” policy, or donate a kidney while alive. We found that providing identifying information about the prospective recipient (whose life was saved by the donation) increased the participants’ willingness to commit to organ donation themselves, donate the organs of a deceased relative, or support a transition to an “opt-out” policy. Conversely, identifying the deceased donor tended to induce thoughts of death rather than about saving lives, resulting in fewer participants willing to donate organs or support measures that facilitated organ donation. A study of online news revealed that identification of the donor is significantly more common than identification of the recipient in the coverage of organ donation cases—with possibly adverse effects on the incidence of organ donations.

Despite the growing number of life-saving organ transplants in recent decades and their increasing success rates, in most countries, not enough organs are recovered from deceased donors to meet the increasing demand (1). In the United States alone, more than 118,000 people are waiting for life-saving organ transplants [as of March 2017, according to the US Department of Health and Human Services (2)]. In many cases, organ transplantation is prevented by the donor’s family’s refusal to permit it—particularly when the deceased had not expressed willingness for organ donation (OD) during their lifetime (3). Organ donation cards, in which people express their agreement to become potential organ donors after their death, are used throughout the Western world—however, the percentage of people who commit to donating their organs after death remains relatively low.

Research suggests that the media are one of the most significant sources of information cited by those seeking to justify their unwillingness to donate their own organs, or those of loved ones (4). Although media coverage of OD issues is largely positive and has become even more positive in the past decade, the details of how the ODs are described may play a crucial role in people’s support—or lack of support—for ODs (5). Specifically, viewers of personal stories that cast ODs in a favorable light—such as TV dramas that make viewers emotionally involved in the narrative—were likely to become organ donors if the drama explicitly encouraged donation (6).

Apart from fictional stories in movies or in television dramas, there are often real cases of OD in the media. In some instances, these stories are presented without any identifying information of the individuals involved; in others, the identity of the donor (such as “a young person recently killed in a road accident”) or the identity of the recipient (e.g., “someone who has been waiting for a kidney for several years”) is given. How might these stories affect readers and their decisions about OD issues? Specifically, we asked which of the above presentations would encourage more people to commit to donating their own organs, or those of a deceased family member, and which would encourage more people to support policies that might increase ODs.

Studies in the past decade have demonstrated that providing identifying information about people in need increases people’s readiness to help (7–10). In particular, when the needs of a single identifiable individual are presented, emotional responses immediately come into play, with higher contributions from the public as a result (10–14). By the same token, when needy individuals are perceived in a negative light, any identifying information about them may increase feelings of anger and blame toward them within the public, resulting in less help being offered (15, 16). This effect is in line with the findings of studies on judgment and decision-making that emotions play a key part in the process of constructing values and preferences (17, 18).

However, the presentation of a victim in need of help, as described above, may be fundamentally different from the presentation of prospective recipients of ODs. When people donate money to aid an identified victim, they believe that their donation will directly help that specific individual whereas, in the case of ODs, the commitment to help is directed toward an unknown future recipient, in the unfortunate event of the donor herself (or a family member of hers) dying. Thus, the presentation of a specific case can be used only by way of illustration, rather than as an actual request for help. Most importantly, when people consider the issue of ODs, they are confronted with the disturbing thought of their own demise, or that of a relative. Such self-focused emotions (such as distress or fear) have been found to play a major role in the initial decision of whether or not to offer help (19).

According to terror management theory (20), prosocial action helps to suppress anxiety-inducing thoughts of death. Thus, people may act prosocially to shield themselves from the looming prospect of their own mortality—inasmuch that, by helping others, they feel more valuable and the world seems more meaningful (21). However, Hirschberger et al. (22) suggest that, when an appeal for help makes the prospect of one’s own death all the more present, people may react by setting it aside and may avoid extending help. For example, in one of their studies, mortality–salience manipulation increased charitable donations but decreased OD card signings (compared with a control condition).

Taking the above research lines together, we suggest that exposure to a specific case of an OD may increase or decrease the salience of one’s own death, depending on who is being identified in the information. When vivid identifying information is provided about the prospective recipient, greater attention is drawn to the possibility of saving a life. Conversely, when detailed information about the donor (a deceased person who has donated his or her organs) is given, thoughts of death predominate, which puts readers or viewers in mind of their own death and may deter them from committing to OD themselves, from donating the organs of a deceased relative, or from supporting policies that may increase ODs.

According to the availability heuristic, people tend to rely on immediate examples that come to mind when evaluating a specific topic (23). Thus, it is important to examine which images of OD cases are more common in the media: an identified donor or an identified recipient? Images from the media are likely to be evoked in people’s minds when they think about the concept of OD. We therefore began with an online investigation that aimed to learn about the way in which OD cases are actually presented in the media. (In this study, we had no intention of judging or criticizing the media. Our only goal was to learn which type of story was more common: a story with an identified donor or a story with an identified recipient.)

An Online Investigation

For information about how OD cases are presented in the popular media, we reviewed all articles on that topic in an 18-mo period (January 1, 2014 through June 1, 2015) on four popular news sites—two of them in Israel (Ynet and Maariv) and two in the United States (USA Today and The Wall Street Journal). The search terms we used were as follows: organ donation, transplantation, transplant, transplanted, donation and kidney, donation and liver, donation and heart, donation and lung/s, donation and pancreas, donation and corneal. Note that we examined only references to ODs from deceased donors; thus, references to ODs from living donors were not included in this analysis.

We found that articles about ODs fell under four categories: general articles on the subject, without reference to a specific case; opinion pieces (also without reference to a specific case); features about a celebrity who had signed an organ donor commitment; and features about individual (noncelebrity) organ donors or recipients. Of these, the last category was the most relevant to us because we were interested in how specific cases of ODs are presented in the media. In total, we reviewed 229 articles (78 in USA Today; 53 in The Wall Street Journal; 56 in Ynet, and 42 in Maariv). Of these, 127 (55.5%) made reference to a specific case of OD: 40 in USA Today, 33 in The Wall Street Journal, 28 in Ynet, and 26 in Maariv. We then examined all of these—specifically, whether the donor or the recipient was identified (e.g., by name or by photograph). We found that there were significantly more articles with identifying details about the donor (58.3%) than about the recipient (11%), about both (18.9%), or about neither (11.8%)—χ² = 76.87, P < 0.001. Moreover, the incidence of articles with identifying information about the donor was similar across all four websites (USA Today, 52.5%, χ² = 17.40, P < 0.001; Wall Street Journal, 51.5%, χ² = 12.45, P < 0.001; Ynet, 60.7%, χ² = 23.14, P < 0.001; Maariv, 73.1%, χ² = 32.15, P < 0.001).

In summary, our brief investigation into the presentation of OD cases in the media revealed that features about specific OD cases predominantly provide identifying information about the donor rather than about the recipient.

Next, we experimentally examined the effect of the presentation of OD cases on people's decisions regarding OD issues by manipulating the identifiability of the donor and the recipient. We expected that providing identifying information about the prospective recipient would increase support for OD issues whereas identifying the deceased donor would decrease support for OD issues by inducing thoughts of death rather than thoughts about saving lives.

Studies Overview

We examined our hypothesis in four studies. In studies 1 to 3, participants read a scenario involving a life-saving OD from one of three perspectives—an identified donor, an identified recipient, or with no identifying information about either party—after which they were asked whether they were willing to sign a bona fide organ donor card (study 1), to consider the donation of a deceased relative's organs (study 2), or to state whether they supported or objected to a change in legal policy that may increase ODs (study 3). In study 4, we manipulated the identifiability of the donor (identified vs. unidentified), as well as the donor's state (a deceased donor vs. a living donor), while examining participants' willingness to donate a kidney. Participants' predominant thoughts about death versus about saving a life were assessed in studies 3 and 4. Ben-Gurion University's institutional review board approved all studies and informed consent was obtained from all participants.

Study 1

Methods.

We recruited 130 undergraduate students who had not signed an organ donation commitment in the past to participate in the study at the end of a class—75% female, mean age 27.07 y, SD = 6.53.* They were randomly assigned to one of three between-subject experimental conditions: an identified donor, an identified recipient, and a control unidentified condition. Participants received a short questionnaire and were asked to read and complete it accordingly, without referring to previously completed pages. It began with a story about a young man who had been killed in a car accident the previous week. The man was a registered organ donor so his parents decided to donate his organs. His kidney was transplanted into the body of another young man, whose life was saved as a result. In the identified donor condition, the name and picture of the deceased donor were given; in the identified recipient condition, the same name and picture were attributed to the organ recipient. In the control condition, no identifying information was given for either person. On the next page, participants were advised that the current study was being carried out in collaboration with the Israeli National Transplant Center (known by its Hebrew acronym, ADI), which maintains a national register of potential organ donors and which participants could join by signing a donor card indicating their willingness to donate their organs after death to save the lives of others. Next, they were presented with an actual ADI donor card and asked whether they were willing to sign it. They could choose between signing, not signing, or declaring that they were already registered (and if so, to provide their ID number for verification). [For ethical reasons, we returned the OD form to the participants at the end of the experiment and told them that they could sign it later, at their convenience—lest their signature be influenced by our priming rather than reflect their initial intention. However, participants did not know about this procedure while making their decision (in any case, this procedure was not expected to interact with our manipulations).]

Results.

Because the DV in this study was willingness to sign the card, those who declared that they had previously signed the card were excluded from further analysis (Study 1, Methods). Of the remaining 130 participants, we examined how many were willing to sign the card. The results showed a significant difference in percentage of signers under the different conditions (χ² = 8.35, P = 0.015). Only 19 participants signed the card during the experiment. Of these, 11 were in the identified recipient condition (27.5%), 5 in the control condition (12.2%), and 3 in the identified donor condition (6.1%). The χ² tests indicated that the percentage of signers in the identified recipient condition was significantly higher than that in the identified donor condition (χ² = 7.59, P = 0.007). The difference between the percentage of signers in the identified recipient condition versus the control condition approached significance (χ² = 2.99, P = 0.07) whereas the two other groups did not significantly differ.

The results of study 1 provide initial evidence for the idea that presentation of OD cases may affect the perceiver's willingness to commit to ODs. In line with our hypothesis, significantly more people were willing to sign the card after reading the case with the identified receiver, than after reading the same case with the identified donor or with no identifying information.

In the next studies, we further examined the influence of the presentation of OD cases on people's decisions regarding organ donation issues. Because the decision about signing an organ donation commitment (used as the DV in study 1) was not relevant to the participants who had already signed the card in the past, there was a potential confound between the identifiability manipulation and the participants' previous decision. To deal with this limitation, in the next three studies, we used different OD decisions that were relevant to all of the participants. In study 2, we examined how the different presentations may influence people's decisions about the donation of the organs of a deceased relative. Because the final decision about OD of the deceased is in the hands of the family, understanding such influences on this decision is of great importance.

Study 2

Methods.

We recruited 164 undergraduate students to complete a short survey individually while working at the library: 50% female, mean age = 26.75 y, SD = 2.13. [Based on power analysis α = .05, a sample of 164 participants was sufficient to detect medium effect size (.30) (24).] Participants were randomly assigned to one of six experimental conditions in a 3 × 2 experimental design of identification (identified donor, identified recipient, control) × information (deceased willing to donate organs known versus no information), as explained below. Participants were provided the same story as in study 1—of a young man who had died in a car accident and whose kidney was transplanted into the body of another young man whose life was saved as a result—under one of the three identification conditions described in study 1. Next, participants were asked to imagine that a close relative of theirs had just died and that the hospital's medical staff were asking their family to consider donating one of his kidneys to save the life of someone waiting for a kidney donation.

The other independent variable manipulated between subjects was information, or lack thereof, of the deceased’s own willingness to donate his organs after death: In one condition, participants were told that he was a registered organ donor whereas, in the other condition, no information was given as to whether he was willing to donate his organs after death.

Participants were then asked to rate their willingness to donate their deceased relative’s kidney, on a seven-category scale ranging from 1 (strongly disagree) to 7 (definitely agree). Finally, they were asked to provide demographic information about themselves and to indicate whether they themselves were registered organ donors (yes/no).

Results.

Results of a 2 × 2 × 2 ANOVA revealed that participants who were told that their relative had indicated that he was willing to donate his organs after death were more amenable to the donation (M = 6.01) than those who were not told this information (M = 5.54), F(1,152) = 4.91, P = 0.028, η_p² = 0.03. Participants who had signed the card themselves in the past were more willing to donate their relative’s organs (M = 6.61) than those who had not made such a commitment (M = 4.94), F(1,152) = 62.65, P < 0.0001, η_p² = 0.29. Finally, the role of identification was also significant, F(2,152) = 3.55, P = 0.03, η_p² = 0.045. In line with the results of study 1, participants in the identified recipient condition expressed a greater willingness to donate the organs of their deceased relative (M = 6.11) than those in the identified donor condition (M = 5.45, P = 0.05). The results of the control group fell in between the two other conditions (M = 5.77) and were not significantly different from either one.

Moreover, as can be seen in Fig. 1, the identification factor had a greater effect on participants who had not signed the card in the past, F(2,57) = 4.54, P = 0.015, η_p² = 0.14, than those who had (F < 1, not significant).

Fig. 1.

Willingness to donate a relative’s organs, as a function of the identification condition and the participants’ own commitment to donating their own organs after death.

The results of study 2 further suggest that providing identifying information about the receiver (rather than about the donor) when presenting OD cases increases support for ODs. The different presentations tend to have a greater impact on people who are not recruited to OD issues in the first place.

Study 3

In this study, we used the same method as in the other two studies: to examine how the presentation of OD cases affects public policy, as opposed to personal decisions. Specifically, the dependent variable in this study was the participants’ willingness to support the transition to an “opt-out” policy in Israel. Research suggests that the percentage of people who choose to become potential organ donors varies greatly between countries, in accordance with how voluntary consent is obtained. In countries with an opt-out policy—that is, where anyone who has not explicitly refused is considered an organ donor after death—consent rates are around 70%, whereas, under an “opt-in” policy (where only those who have given explicit consent are donors), only 10 to 40% of the population agree to being organ donors (25, 26). Although an opt-out policy does not guarantee higher per capita rates of organ recovery (because next-of-kin may choose to veto the OD at the critical moment) (27), a legal condition that mandates donation by default tends to increase the incidence of that option (28).

To learn more about the mechanism behind the effect, in this study, we also asked participants to indicate how much they were thinking about saving a life as opposed to thinking about death (their own death or someone who has died) when considering their response to the policy decision.

Methods.

Two hundred thirty-nine undergraduate students completed a short questionnaire voluntarily: 68% females, mean age = 25.31 y, SD = 3.17. [Based on χ² power analysis α = .05, df = 2, a sample of 240 participants was sufficient to detect medium effect size (.03) (24).] Participants were randomly assigned to one of the three experimental conditions described in study 1 (identified donor, identified recipient, and control). After reading the scenario, participants in all conditions were provided with a short explanation about ADI (the Israeli National Transplant Center) and the national organ donors register. They were then advised that, under the existing policy in Israel, citizens are not registered organ donors after death unless they expressly registered as such; however, in some countries, all citizens are considered potential organ donors unless they expressly opted out. Participants were then told that studies consistently show that the percentage of organ donors is significantly higher under the opt-out policy than in the opt-in policy. Finally, they read that, in the light of these studies, there is a proposal to change the policy in Israel from an opt-in to an opt-out policy. Participants were then asked to express their own opinion about such a change, by choosing between keeping the current opt-in policy or changing it to an opt-out one. (The order of these two options was varied between participants.)

Next, they were asked to note what their prominent thought was when they were making their choice: about someone who had died; about someone whose life was saved; or about their own death. Finally, they were asked to provide demographic information about themselves and to indicate whether they themselves were registered organ donors.

Results.

Overall, 147 participants expressed support for a transition to an opt-out policy (61%). Results of χ² tests indicated that the percentage of participants who supported the opt-out policy significantly differed under the three conditions (χ² = 7.91, P = 0.019). Specifically, the percentage of participants who supported the opt-out policy under the identified recipient condition (73.1%) was significantly higher than that of proponents of that policy in the identified donor condition (51.3%; χ² = 7.88, P = 0.004). The difference between the percentage of proponents in the identified recipient condition versus that in the control condition (60.2%) approached significance (χ² = 2.97, P = 0.09) whereas the numbers in the identified donor condition and the control (unidentified) condition were not significantly different (Table 1).

Table 1.

Number and percentage of participants who supported the opt-in (vs. opt-out) policy under the three experimental conditions

Identification	Policy		Total
	Opt-in	Opt-out
Donor
Count	38	40	78
% within identification	48.7	51.3	100.0
Recipient
Count	21	57	78
% within identification	26.9	73.1	100.0
No identification
Count	33	50	83
% within identification	39.8	60.2	100.0
Total
Count	92	147	239
% within identification	38.5	61.5	100.0

One hundred thirty-six (56.7%) of the participants reported that they had signed an organ donor card in the past. Not surprisingly, these participants expressed greater support for a transition to the opt-out policy (73.5%) than those who had not signed such a card (45.6%; χ² = 19.27, P = 0.001). However, the identifiability manipulation seems to have had a similar effect on signers and nonsigners alike: The percentage of participants in favor of a transition to an opt-out policy in both the identified donor and the identified recipient condition was similar for signers (χ² = 2.68, P = 0.09) and nonsigners (χ² = 4.63, P = 0.05) alike.

Next, we examined what the participants were thinking of while answering the questionnaire. To examine the relative dominance of “thoughts of death” versus “life-saving thoughts” under the various conditions, we grouped together the two death-related concepts (about someone who has died; about my own death) under the category “thoughts of death.” (The percentage of participants who chose one of the two death-related thoughts did not significantly differ between conditions.) The percentages of participants who focused on thoughts of death under the identified donor condition (52.6%) were higher than percentages of participants who focused on thoughts of death under the identified recipient condition (40%) and higher than under the control condition (40.2%). However, these differences did not reach significance (χ² = 2.43, P = 0.06 and χ² = 2.44, P = 0.059, respectively, in a one-tailed test).

Finally, we examined the relationship between participants’ policy decisions and their thoughts while answering the questionnaire. The results of χ² tests significantly showed that participants who supported the transition to the opt-out policy were more likely to focus on life-saving thoughts (64%) than on death (36%) whereas the opposite was true for participants in favor of maintaining the opt-in policy (only 43% focused on life-saving vs. 57% on death; χ² = 9.99, P = 0.002).

The results of study 3 demonstrate how the presentation of OD cases can affect people’s policy decisions in that domain. Adding identifying information about the deceased who opted to donate his or her organs after death to save others resulted in fewer participants supporting the opt-out policy compared with the percentage of participants who supported this transition when adding identifying information about the recipient who was saved by OD. Moreover, the presentation of an identified donor tended to evoke more death-related thoughts than life-saving ones. Death-related thoughts were significantly correlated with a decrease in people’s willingness to support a transition to an opt-out policy.

To examine our hypothesis more directly (i.e., that the presentation of an identified deceased donor decreases willingness to donate because it tends to increase thoughts about death), we manipulated in study 4 the status of the donor (living or deceased) and examined how identifiability affected OD decisions in each case.

Study 4

Methods.

One hundred twenty-six undergraduate students participated in the study voluntarily while working individually at the library. They all read a short story about a young man who donated his kidney and saved the life of another young man who urgently needed kidney transplantation. Participants were randomly assigned to one of four experimental conditions of a 2 × 2 design, manipulating two variables: the status of the donor—a living donor (a young man who decided to donate his kidney while alive for altruistic reasons) or a deceased donor (a young man who died in a car accident and was a registered organ donor)—and the identifiability of the donor—an identified condition, in which the name and the picture of the donor were provided, or an unidentified condition, in which no identifying information was provided. After reading the story, participants were asked to rate the extent to which they were willing to donate a kidney to a sick stranger, given that they were medically suitable for the transplantation on a seven-degree scale.

Next, as in study 3, they were asked to note what their prominent thought was when they were making their choice: about someone who had died; about someone whose life was saved; or about their own death. Finally, they were asked to provide demographic information about themselves and to indicate whether they themselves were registered organ donors.

Results.

Participants' willingness to donate a kidney was analyzed in a 2 × 2 ANOVA. Results revealed no significant main effects. However, the interaction between the donor's status and identification was significant, F(1,122) = 4.63, P = 0.33, η_p² = 0.037. As can be seen in Fig. 2, participants in the deceased donor condition were more willing to donate a kidney in the unidentified condition (M = 4.12) than in the identified condition (M = 3.32), t(62) = 1.84, P = 0.035, one-tailed. However, in the living donor condition, willingness to donate a kidney did not significantly differ under the two identification conditions and tended to be higher when the donor was identified (M = 3.80) than when he was not identified (M = 3.28), t(60) = –1.20, P = 0.23.

Fig. 2.

Willingness to donate a kidney as a function of the donor's status (deceased vs. living) and identification (identified vs. unidentified).

Next, we examined what the participants were thinking of while answering the questionnaire. As in study 3, to examine the relative dominance of thoughts of death versus life-saving thoughts under the various conditions, we grouped together the two death-related concepts (about someone who has died; and about my own death) under the category “thoughts of death.” Looking at the deceased donor condition, the percentages of participants who focused on thoughts of death under the identified donor condition (51.7%) were higher than percentages of participants who focused on thoughts of death under the unidentified donor condition (21.2%; χ² = 6.28, P = 0.017). However, under the living donor condition, only one participant (in the identified donor condition) reported being focused on thoughts of death (χ² = 1.02, P = 0.50).

The results of study 4 directly show that identification of an organ donor decreases willingness for OD only when the donor is a deceased person, a presentation that tends to raise thoughts about death (rather than about saving a life), but not when a living donor is described. In the case of a living donor, participants' thoughts were overwhelmingly focused about saving a life (rather than about death). These findings support our hypothesis that identification of a deceased donor decrease willingness for OD due to the increase in thoughts about death.

Discussion

Families whose loved ones have died and who have decided to donate their organs are usually more eager to publish their story, along with many details about the deceased, and his or her life and death. They may even see the publicity surrounding the donation as a means of commemorating the deceased and giving meaning to his or her death (3). In contrast, the recipients of ODs usually prefer to remain anonymous. As a result, we are more likely to hear or read personal stories that identify the organ donors than the organ recipients. Our study of online news revealed that news items or features about ODs that identify the donor are indeed significantly more common than those that identify the recipient. As mentioned earlier, according to the availability heuristic, people tend to rely on immediate examples that come to mind when evaluating a specific topic (23) so such stories are more “front of mind” than stories that identify organ recipients when we think about the issue of ODs.

The results of the four studies we have presented suggest that detailed descriptions of deceased donors do not motivate people to support OD issues. Rather, our findings suggest that, after reading such stories, people are less likely to commit to donate their organs after death, to consider donating the organs of a deceased loved one, to support a public policy that might promote ODs, and to donate a kidney while alive. In contrast, identification of a prospective recipient of OD makes people more favorably inclined toward ODs. Participants in study 1 who were told about an identified organ recipient were more likely to sign an OD card than those who had read a version of the same story that identified only the donor, or provided no identifying details about either one. Similarly, in study 2, participants were more willing to donate the organs of a deceased loved one after reading about an identified recipient (compared with reading about an identified donor). Finally, in study 3, participants were more likely to support a transition to an opt-out policy after reading about a successful OD procedure in which the recipient was identified than if they had read the same story with an identified donor or with no identifying information at all.

The results of studies 3 and 4 provide some insight into the mechanism behind this effect. Although participants in study 3 who read a story about OD with an identified recipient or with no identifying information tended to focus on the idea of saving a life, those who read a variation of the same story with identifying details about the donor only tended to be more occupied by thoughts of death. Furthermore, thoughts about saving a life were related to the support of opt-out policies whereas the opposite was true for people who were induced into thinking about death. The results of study 4 strengthen this notion by demonstrating that identification of an organ donor decreases willingness for OD only when death is involved and the donor is dead, but not when a living donor is described.

Research on the identifiable victim effect typically examines situations in which the prospective recipient of help is identified (e.g., refs. 8, 12, and 13). Other studies have examined situations in which the identified victim is one of many who are in need (e.g., refs. 29 and 30). The present study demonstrates that identification of someone who is neither the prospective recipient of help nor in urgent need may also affect people’s decisions. The instances that participants in our studies read about involved organ transplants that already had taken place and individuals who had been saved—thus, they could serve to illustrate the issue only in general terms. Further research is needed to examine this unique phenomenon in other prosocial decisions. Similarly, our findings suggest that, when an instance is described, the role of the identified characters can have a significant bearing on the outcome. In particular, identification of characters seems to accentuate their role and what they represent: When they represent a life saved, the effect is different from when they represent someone who has died.

Besides this theoretical contribution, our research suggests practical implications for efforts to promote ODs. For example, recruiting people whose lives were saved by OD, identifying them by name, and telling their story may increase media coverage about such individuals and spur members of the public to think about saving lives when reading about ODs and to view ODs in a favorable light.

Research is needed to examine organ transplant recipients' willingness to tell their story as a function of the type of the transplanted organ (for example, it might be that kidney transplant recipients are more willing to tell their story than heart transplant recipients). Insights from such a study may help in efforts to recruit OD recipients to tell their stories, suggesting which recipients are more likely to cooperate. In general, creating a more optimistic image of OD cases may have a beneficial effect on public opinion on this issue, increase the number of potential organ donors and supporters, and possibly help save the lives of those waiting for organ transplants.