Media and the Kindertelefoon:
The impact of news items about suicide on
conversations to the helpline
Master Thesis
Author: Kèren Uchibori (5837979) University of Amsterdam
Amsterdam Business School Faculty of Economics and Business
MSc. In Business Administration – Marketing track
August 24, 2015
Primary supervisor : Prof. dr. W.M. van Dolen Secondary supervisor : Dr. M.J. Soomer
2
Statement of originality
This document is written by Student Kèren Uchibori who declares to take full responsibility for the contents of this document.
I declare that the text and work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Table of contents
STATEMENT OF ORIGINALITY ... 2
ABSTRACT ... 6
1. INTRODUCTION AND RESEARCH QUESTION ... 7
1.1. RESEARCH QUESTION ... 10
2. THEORETICAL FRAMEWORK AND HYPOTHESIS DEVELOPMENT... 12
2.1. THE WERTHER EFFECT: THE CONTAGION EFFECT IN SUICIDES ... 12
2.1.1. CHARACTERISTICS OF NEWS ITEMS LEADING TO SUICIDAL CONTAGION ... 19
2.2. THE PAPAGENO EFFECT: THE PREVENTIVE EFFECT OF THE MEDIA ON SUICIDES ... 24
2.2.1. TRIGGERS FOR DISCUSSING SUICIDAL THOUGHTS ... 27
3. RESEARCH METHOD ... 30
3.1. SELECTION OF NEWSPAPER ARTICLES ON CHILD SUICIDE ... 30
3.1.1. DESCRIPTIVES OF THE DATASET: NEWSPAPER ARTICLES ... 35
3.2. SELECTION OF TELEVISION BROADCASTS ON CHILD SUICIDE ... 37
3.2.1. DESCRIPTIVES OF THE DATASET: TV BROADCASTS ... 38
3.3. DATABASE KINDERTELEFOON ... 39
3.3.1. BACKGROUND ON THE KINDERTELEFOON ... 39
3.3.2. TOPICS OF CONVERSATIONS ... 40
3.3.2.1. SUBTOPICS ... 42
3.3.3. NUMBER OF CONTACTS (TELEPHONE AND CHAT CONVERSATIONS COMBINED) ... 43
3.3.4. THE TYPE OF CONVERSATION ... 44
3.3.5. MEDIUM (TELEPHONE CALLS VERSUS CHATS) ... 44
3.3.6. AGE ... 45
3.3.7. GENDER ... 45
3.3.8. DAY OF THE WEEK ... 46
3.3.9. DURATION OF CONVERSATION ... 46
4. RESULTS ... 47
5. DISCUSSION AND CONCLUSION ... 55
5.1. MANAGERIAL IMPLICATIONS... 57
5.2. LIMITATIONS OF THE PRESENT RESEARCH AND SUGGESTIONS FOR FUTURE RESEARCH ... 59
6. REFERENCES ... 63
APPENDIX I. OVERVIEW OF NEWS ITEMS ABOUT SUICIDE (TELEVISION AND NEWSPAPERS), 1994-2013 ... 67
4
APPENDIX II. TABLES AND GRAPHS ... 80
TABLE 1.NUMBER OF NEWSPAPER ITEMS ABOUT CHILD SUICIDE VS. NUMBER OF TV BROADCASTS ON CHILD SUICIDE, 1994-2013 AGGREGATED ... 80
TABLE 2.NEWSPAPER ITEMS ABOUT CHILD SUICIDE; GENDER OF THE CHILD THAT COMMITTED SUICIDE, 1994-2013 AGGREGATED ... 80
TABLE 3.NEWSPAPER ITEMS ABOUT CHILD SUICIDE; NUMBER OF NATIONAL VS REGIONAL NEWSPAPERS, 1994-2013 AGGREGATED ... 80
TABLE 4.COVERAGE OF NEWSPAPERS PER TITLE, PER YEAR.1994-2013 ... 82
TABLE 5.COVERAGE OF NEWSPAPERS PER TITLE.FIXED (AVERAGE) NUMBER PER TITLE ... 83
TABLE 6.NEWSPAPER ITEMS ABOUT CHILD SUICIDE; COUNTRY OF THE SUICIDE VICTIM,1994-2013 AGGREGATED... 83
TABLE 7.NEWSPAPER ITEMS ABOUT CHILD SUICIDE; AGE OF THE SUICIDE VICTIM,1994-2013 AGGREGATED ... 83
TABLE 8.NEWSPAPER ITEMS ABOUT CHILD SUICIDE; REASON OF THE SUICIDE AS MENTIONED IN THE ARTICLE (IF APPLICABLE),1994-2013 AGGREGATED ... 84
TABLE 9.NEWSPAPER ITEMS ABOUT CHILD SUICIDE;CONTEXT CATEGORIES (DERIVED FROM PIRKIS, ET AL. 2007),1994-2013 AGGREGATED ... 85
TABLE 10.NEWSPAPER ITEMS ABOUT CHILD SUICIDE;CONTAGION CATEGORIES (DERIVED FROM CARROLL ET AL.,1994),1994-2013 AGGREGATED ... 85
TABLE 11.TV BROADCASTS ON CHILD SUICIDE; GENDER,1994-2013 AGGREGATED ... 85
TABLE 12.TV BROADCASTS ON CHILD SUICIDE; COUNTRY,1994-2013 AGGREGATED ... 86
TABLE 13.TV BROADCASTS ON CHILD SUICIDE; AGE,1994-2013, AGGREGATED ... 86
TABLE 14.TV BROADCASTS ON CHILD SUICIDE;CONTEXT CATEGORIES (DERIVED FROM PIRKIS ET AL. 2007),1994-2013 AGGREGATED ... 86
TABLE 15.TV BROADCASTS ON CHILD SUICIDE;CONTAGION (DERIVED FROM CARROLL ET AL.1994), 1994-2013 AGGREGATED ... 87
TABLE 16.MOST DISCUSSED MAIN TOPICS IN CONVERSATIONS TO THE KINDERTELEFOON IN 2013; TELEPHONE VS. CHAT ... 87
TABLE 17.KINDERTELEFOON;LIST OF MAIN TOPICS AND SUBTOPICS USED BY THE KINDERTELEFOON TO CATEGORIZE CONVERSATIONS... 88
CHART/GRAPH 1.KINDERTELEFOON;MAIN TOPICS DISCUSSED,1994-2013 AGGREGATED ... 89
TABLE 18.KINDERTELEFOON;FREQUENCY OF MAIN TOPICS DISCUSSED (SERIOUS CONVERSATIONS), 1994-2013 AGGREGATED ... 89
TABLE 19.KINDERTELEFOON;NUMBER OF CONTACTS PER SUBTOPIC (SERIOUS CONVERSATIONS). 1994-2013 AGGREGATED ... 89
TABLE 20.KINDERTELEFOON;SHARE OF CONVERSATIONS ABOUT SUICIDE PER YEAR (SERIOUS CONVERSATIONS),1994-2013 ... 90
CHART/GRAPH 2.KINDERTELEFOON;SHARE OF CONVERSATIONS ABOUT SUICIDE PER YEAR (SERIOUS CONVERSATIONS),1994-2013(2008 EXCLUDED) ... 90
TABLE 21.KINDERTELEFOON; REGISTERED VARIABLES PER CONVERSATION ... 91
CHART/GRAPH 3.KINDERTELEFOON; AMOUNT OF CONTACTS PER YEAR,1994-2013 ... 91
TABLE 22.KINDERTELEFOON; EVENTS CAUSING EXPLAINED INCREASES/DECREASES IN AMOUNT OF CONVERSATIONS ... 92
TABLE 23.KINDERTELEFOON;OVERVIEW OF THE AMOUNT OF EACH TYPE OF CONVERSATION, PER YEAR, 1994-2013 ... 93
TABLE 24.KINDERTELEFOON;AMOUNT OF CONVERSATIONS:TELEPHONE VS CHAT, PER YEAR,1994-2013 ... 93
5 CHART/GRAPH 4.KINDERTELEFOON;AMOUNT OF SERIOUS CONVERSATIONS:TELEPHONE VS CHAT, PER YEAR,1994-2013 ... 94
CHART/GRAPH 5.KINDERTELEFOON: AMOUNT OF (SERIOUS) CONVERSATION ON THE SUBTOPIC OF SUICIDE, PER YEAR,1994-2013... 94
CHART/GRAPH 6.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS,AGE,1994-2013 AGGREGATED ... 94
TABLE 25.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS,AGE (FREQUENCY),1994-2013
AGGREGATED... 95
CHART/GRAPH 7.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS,AGE, CHAT VS TELEPHONE,
1994-2013 AGGREGATED ... 95
CHART/GRAPH 8.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS ON SUICIDE,AGE, PER YEAR,
1994-2013 ... 95 CHART/GRAPH 9.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS ON SUICIDE,AGE (IN %),
1994-2013 AGGREGATED ... 95
CHART/GRAPH 10.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS, GENDER, PER YEAR,1994-2013 .. 96
CHART/GRAPH 11.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS BY TELEPHONE, GENDER.
1994-2013, 2008 EXCLUDED ... 96
CHART/GRAPH 12.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS BY CHAT, GENDER, PER YEAR,
1994-2013 ... 96 CHART/GRAPH 13.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS ON SUICIDE.GENDER.
PER YEAR,1994-2013, EXCLUDING 2008 ... 97
TABLE 26.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS.DAY OF THE WEEK.1994-2013
AGGREGATED... 97
CHART/GRAPH 14.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS.DAY OF THE WEEK.
1994-2013 AGGREGATED (IN %) ... 97
CHART/GRAPH 15.KINDERTELEFOON;ALL CONVERSATIONS ABOUT SUICIDE (SERIOUS CONVERSATIONS).
DAY OF THE WEEK. PER YEAR,1994-2013(2008 EXCLUDED) ... 98
CHART/GRAPH 16.KINDERTELEFOON;ALL CONVERSATIONS ABOUT SUICIDE (SERIOUS CONVERSATIONS).
WEEK NUMBER.PER YEAR,1994-2013(2008 EXCLUDED) ... 99
TABLE 27.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS.DURATION.TELEPHONE VS CHAT.
1994-2013 AGGREGATED ... 100
CHART/GRAPH 17.KINDERTELEFOON;ALL SERIOUS CONVERSATIONS.DURATION.TELEPHONE VS CHAT.
1994-2013 AGGREGATED ... 100
CHART/GRAPH 18.KINDERTELEFOON;ALL CONVERSATIONS ABOUT SUICIDE (SERIOUS CONVERSATIONS).
6
Abstract
Objective: This research examined the influence of news items (both television broadcasts
and newspaper reports) in the Netherlands about child suicide on the amount of conversations to the Kindertelefoon (a helpline for children, in the Netherlands), specifically on the topic of suicide.
Methods: Regression analysis was conducted on the relationship between (the content of)
press coverage on suicides by children (on a weekly basis) and conversations to the Kindertelefoon (on a weekly basis) from January 1994 to September 2013.
Results: The results show that a fixed coverage figure per newspaper title, which served as an
index for the amount of people that came in contact with announcement about child suicide, explained 9.90% (p < .001) of the variance in conversations to the Kindertelefoon about suicide. This effect was mainly observable in chat conversations, rather than in telephone conversations.
Conclusion: As an significant effect was found of newspaper items about child suicide on the
amount of chat conversations to the Kindertelefoon about suicide, it is advisable to inform employees of the Kindertelefoon of the effect found in this research, and for the Kindertelefoon to monitor current events (reports about child suicide) and informing employees in case of these events. This way, they can be even better prepared for answering (difficult) questions about suicide by placing the questions in the right context.
7
1. Introduction and research question
In The Netherlands suicide has been the second most common cause of unnatural death among children and teenagers, both boys and girls, for over the past decades. With fifty-eight cases in 2013 it has reached the highest point since 1982. Accidents, including for the most part traffic accidents, are the primary cause of unnatural death among children and teenagers. Most other countries show the same trend (Becker & Schmidt, 2005). Teenagers are confronted with stressors and social pressure which may make them consider suicide as a solution to their problem(s) (Elliot & Feldman, 1990).
The rate of suicides may be small, but the impact on society is bigger than this ratio might suggest. Suicides are deaths that are avoidable to a certain extent. Usually a period of psychological problems precedes the suicide act, during which also help can be offered. After a suicide, the relatives and friends of the person that committed suicide also suffer severely from the bereavement which is why Statistics Netherlands calls suicide an indicator of psychological well-being in a country (Centraal Bureau voor de Statistiek (2010, 2014).
Research shows that children are susceptible to both positive and negative influences from peers and others. Information coming from the media can also have an influence on the feelings and behavior of children (Portzky, Audenaert & Heeringen, 2005). More specifically, as is further described in the next chapter, research shows that the way the media report on suicide cases, can influence subsequent suicidal behavior (e.g. Stack, 2005). For example, when news items on a suicide case are of a sensational nature or about a celebrity suicide for example, an increase in suicides is found thereafter. Several authors (e.g. Etzersdorfer & Sonneck, 1998; Phillips, 1974) suggest that vulnerable people are susceptible for suggestions regarding suicide, including the method used, and that media reports may make them identify
8 themselves with the person described in the article, and they may think committing suicide is a solution to their problems. This phenomenon is called ‘suicidal contagion’. Research suggests that suicidal contagion is particularly acute among vulnerable teenagers, who may copy behavior of people they can identify with (Baume, et al., 1997; Chen, et al. 2010; Gould, et al., 2003; Stack, 2003).
On the other hand, when media reports on suicide are concise and factual, and include telephone numbers of a helpline, an increase in help-seeking behavior is found (Niederkrotenthaler, et al., 2010). However, this may not necessarily lead to a decrease in suicides. Children have various options for obtaining help; informal (family and friends) and formal (professionals). One of the formal options to seek help from is to use a helpline (Potter & Hepburn, 2003). This can either be by telephone or online (via chat). These helplines are an easy accessible way of help-seeking, as they usually can be contacted anonymously, without parental consent, and without costs (Tylee, Haller, Graham, Churchill, & Sanci, 2007). In The Netherlands, the Kindertelefoon offers counseling services to children, both via telephone and chat. All conversations are treated anonymously, but certain details (including the topic) are registered (Kindertelefoon, 2013).
In The Netherlands, several studies were performed on suicide in general but only one research on suicidal contagion was found (Köpping, Ganzeboom, & Swanborn, 1989). This is a replicating study of Phillips’ research (Phillips, 1974). No research on the influence of media attention about suicide on help-seeking behavior was found.
This thesis covers the research on the influence of press coverage (both newspapers and TV broadcasts) about suicides that are committed by children on the number of conversations to the Kindertelefoon, especially on the topic of suicide. It focuses only on children, as they
9 are said to be more susceptible to media influences and suicidal contagion (Baume, et al., 1997; Chen, et al. 2010; Gould, et al., 2003; Pearson Gibson & Range, 1994; Stack, 2003), and as they are the target group of the Kindertelefoon.
The contribution of the present research is in the field of Services Marketing, more specifically Consumer Behavior. The children contacting the Kindertelefoon are the helpline’s users/clients. This research builds on existing literature and aims to contribute to it by getting a better understanding of the effect of media reports on suicides by children on help-seeking behavior, specifically to the Kindertelefoon and on how and when a service provider can adjust its services to current affairs if necessary. This research aims to contribute to the work of the Kindertelefoon by creating awareness among employees of effects of current affairs that become top op mind and trigger children to contact the Kindertelefoon, in order to help children in distress. The volunteers of the Kindertelefoon find the subject of suicide one of the most difficult to deal with (Kindertelefoon, 2007), which is one of the reasons why this topic is looked into further in the present research. Additionally, by analyzing if media attention triggers or does not trigger conversations on a certain topic, suicide in this case, the Kindertelefoon can adjust its service if needed. For example, if the research shows peaks in help-seeking behavior after media attention on suicide, the Kindertelefoon may want to register this media attention and adjust their staffing accordingly in order to provide a better service to its users/clients; additional volunteers and/or specific attention to the possibility of receiving calls/chats about suicide. Another example may be when the research shows that a certain day (e.g. Wednesday afternoon, when most primary schools close early) shows an increase in conversations on suicide of the week, the Kindertelefoon may adjust staffing after
10 media attention on suicide. This could serve as an example to other service providers; a specific service following a specific event.
1.1.
Research question
The research question related to the information mentioned before and additional reasoning is as follows:
“To what extent do news items, both in newspapers (national and regional) and on
television in The Netherlands about suicides, including murder-suicides, by children until 18 years, influence the amount of calls and chat conversations made to the Kindertelefoon by children on the topic of ‘suicide’ in the period January 1994 – September 2013?”
This period has been chosen because there is a lot of detailed data available from the Kindertelefoon on during those years on a daily basis.
In order to answer the research question, the following questions need to be answered first:
A. What is known in the literature about suicidal contagion?
A.1. What characteristics of news items trigger suicidal contagion? A.2. What characteristics of news items trigger help-seeking behavior?
B. What articles have been published in Dutch national and regional newspapers and what television broadcasts were aired about suicides by children until 18 years old in the period January 1994 to December 2013?
B.1. What are the characteristics of these articles/broadcasts?
C. What are the details of the calls/chat conversations to the Kindertelefoon in the period January 1994 to December 2013?
11 The next chapter will cover the theoretical framework (question A). Chapter 3 covers the research method (question B). The data needed to answer question C were provided by the Kindertelefoon, for which a confidentiality agreement was signed by the author. The dataset is not included in this document. Descriptives of this set are provided in chapter 3.
12
2. Theoretical framework and hypothesis development
Suicide is a global public health concern, claiming more than 1 million lives globally each year (World Health Organization, 2014). People commit suicide for a variety of reasons different reasons, but there is evidence that the media also have a powerful influence over people in distress.
The influence of the media can be both negative, leading to more suicides, and positive, leading people to refrain from committing suicide and seeking help instead.
Both effects will be discussed in this chapter. This chapter is started with a description of the
Contagion Effect in Suicide, followed by a description of specific factors in newspaper articles
that may increase the likelihood of a contagion effect, such a sensational description of a suicide case in the media or the reporting of a suicide by a celebrity.
Next, the preventive effect of the way of reporting of suicides by the media is described. For example by the introduction of media guidelines on responsible reporting and also by including contact details of helplines in the articles.
2.1.
The Werther effect: the contagion effect in suicides
Over two centuries ago, in 1774, German writer and statesman Goethe wrote “Die Leiden des jungen Werthers” (translated: The sorrows of young Werther). The main character in this book, a young artist named Werther, committed suicide. It was a popular book and it was said that people in many European countries copied Werther, by not only dressing like him, but moreover by committing suicide. As a result, the book got banned in several regions in Europe (Gould, 2001; Gray, 1967; D.P. Phillips, 1974).
13 Phillips (1974) was one of the first to investigate this so-called contagion effect in suicides, which he named the ‘Werther effect’ after the main character in Goethe’s book. He defined this effect as ‘the increase in number of suicides after the story of a suicide is publicized in the newspapers’.
In his first study on this subject, Phillips (1974) found a significant relationship between publicized suicide stories on the front-page of major U.S. and British newspapers, and the number of actual suicides in the United States and Great Britain respectively in the period from 1947 to 1968. By conducting content analysis on the headlines of major newspapers he showed that the number of actual suicides increased nationwide and sometimes even internationally in the month after publication of a newspaper article on a suicide story. Phillips (1974) found that the more media attention that was given to a suicide story, the larger the rise in suicides thereafter. The Werther Effect was also influenced by geographical location, meaning that if a suicide story was mainly publicized in New York, the monthly suicide number increased in New York more so than in other parts of the United States (Phillips, 1974). Phillips (1974) also found an increase of 12 per cent in the suicide rate in the month after the suicide of celebrity Marilyn Monroe. In later studies, the same author (1977, 1978, 1979, 1980) controlled for trends and seasonal influences in suicides and consistently showed a significant relationship between the amount of publicity of a suicide case and the increase in suicides thereafter by two to four per cent.
Ever since Philips’ studies were published, various authors have conducted research on imitative suicidal behavior after media coverage of either real or fictional suicide stories, and research was extended to include suicides broadcasted on television and online (which is
14 called ‘cyber suicide’ by Baume, Cantor and Rolfe (1997) next to newspaper coverage, resulting in a vast amount of literature on this subject.
In the literature the Werther Effect is also referred to as the contagion/copycat effect in
suicides, (Etzersdorfer & Sonneck, 1998; Gould, 2001; O'Carroll & Potter, 1994), and the suggestion-imitation model (Hittner, 2005). In this text these terms will be used
interchangeably.
Authors suggested that people that are vulnerable and are considering suicide are susceptible for ideas about suicide and that a newspaper article on a suicide case may make them identify themselves with the person described in the article, which may then work as a trigger for committing suicide, or they may use the described method to commit suicide (Etzersdorfer & Sonneck, 1998).
The definition of the Werther Effect was extended and now includes the effect of television reports and other media of suicide cases on the amount of suicides as well. Based on literature, Pirkis and Blood (2010) updated and broadened the definition to ‘the relationship between media portrayals of suicide and imitation acts, including completed suicides, attempted suicides and suicidal thoughts.’
Most of the studies have proven to a certain extent that newspaper coverage on suicides may lead to more suicides. See Pirkis and Blood (2010) for a comprehensive overview. Studies were not only performed in the United States, but also in other countries.
In The Netherlands, Köpping, Ganzeboom & Swanborn (1989) performed content analysis on three newspapers with a national range of distribution in The Netherlands over a period of 30 years and of several different newspapers. Only the length of an article was found to have
15 a significant effect; longer newspaper articles led to a small increase in suicides after they were publicized, but the majority of other relationships such as a reference to suicide in the headline, a photograph included in the article or the mentioning that it was a well-considered act were not significant (Kopping et al., 1989).
However, most other studies on the Werther Effect from newspapers, from the United States from different time periods and various countries including Hungary (e.g. Fekete & Macsai, 1990), Germany (e.g. Jonas, 1992), Japan (e.g. Stack, 1996), Austria (e.g. Etzersdorfer, Voraeck & Sonneck (2004) did show that a copycat effect exists to a certain extent.
Both newspaper reports and television reports show that younger age groups are more susceptible to the Werther effect than older people (Baume et al., 1997; Chen et al., 2010; Fekete et al., 2001; Gould, Jamieson, & Romer, 2003; D.P. Phillips, 1974; David P Phillips & Carstensen, 1986; A Schmidtke & Hafner, 1989; S. Stack, 2003).
By performing a logistic regression analysis on over 40 studies on the Werther effect, Stack (2003) showed the copycat effect was stronger (4.03 times more likely) in studies that were based on real cases of suicide in comparison to fictional stories (Stack, 2003). According to social learning theory this is because one is more likely to relate to a real person than a fictional person (Bandura, 1977; Pirkis, 2009).
The present study does not cover the effect of news items on actual suicides, but reasoned is that the amount of calls/chat conversations to the Kindertelefoon on certain topics increases, based on the content of news items.
16
The following hypotheses were derived from the theory and subsequent reasoning:
Hypothesis 1: News items in newspapers about child suicide will positively influence the amount of conversations to the Kindertelefoon on the topic of suicide.
Hypothesis 1a: News items in newspapers about child suicide will positively influence the amount of telephone conversations to the Kindertelefoon on the topic of suicide.
Hypothesis 1b: News items in newspapers about child suicide will positively influence the amount of chat conversations to the Kindertelefoon on the topic of suicide.
Studies on television reports (e.g. Gould & Shaffer, 1986; Hittner, 2005; Phillips & Carstensen, 1986; Schmidtke & Hafner, 1988), both non-fictional and fictional, on suicides did also find a contagion effect, however this was not as strong as the effect of newspaper reports on suicides. In a meta-analysis Stack (2003) found that studies on the copycat effect of television stories were 82% less likely to find this effect than studies on newspaper articles on suicides. Authors argue that newspaper articles have a stronger impact than television reports, as newspapers can be kept and read more than once (Pirkis & Blood, 2001a, 2001b; Stack, 2003).
Hypothesis 2: News items on television about child suicide will positively influence the amount of conversations to the Kindertelefoon on the topic of suicide.
Hypothesis 2a: News items on television on child suicide will positively influence the amount of telephone conversations to the Kindertelefoon on the topic of suicide.
Hypothesis 2b: News items on television on child suicide will positively influence the amount of chat conversations to the Kindertelefoon on the topic of suicide.
17 So far, not many studies have been carried out on the influence of the internet on suicidal behavior. There are chat rooms for example, where suicide methods are discussed and suicide is even promoted, and there are also chat rooms where information on helplines is exchanged (Becker & Schmidt, 2005). The studies that have been performed (e.g. Baume, Cantor & Rolfe, 1997; Becker & Schmidt, 2005) show some evidence to suggest a negative influence of this medium on suicidal behavior, but other authors claim more research is required to confirm a causal relationship (Pirkis & Blood, 2010).
Considering the subject of suicide, experimental studies are not easy to carry out on this topic. Most studies started after an increase in suicides occurred, when researchers tried to find a suspected Werther effect. However, some authors have performed (quasi-) experiments to contribute to the knowledge on the Werther effect; Pearson Gibson and Range (1991) argued that in earlier studies one could not be sure if the person that had committed suicide had actually read a newspaper article or saw a television broadcast on a suicide case, so the increase in the rate of rate of suicides could have been caused by other reasons, rather than by the Werther Effect. Therefore, they designed and performed a quasi-experiment. High-school students were asked to read one out of four stories on a teenager in distress. Some stories included information that the teen knew someone with similar problems who asked for professional help and other stories mentioned that the teenager had committed suicide. The students were then asked what they thought what the teenager might do. Results showed that the students thought contagion would occur for both suicide and asking for help (Pearson Gibson & Range, 1991). They also showed that girls tended to be more open to seeking help, while boys were slightly more responsive to suicide. However, people may
18 respond in a different way in a real-life situation than they do to a fictitious story (Gibson & Range, 1991).
Hawton, et al. (1999) also conducted an experiment. In a direct assessment, they studied individual patients and found an increase of suicide attempts by an overdose of paracetamol after a television broadcast of the soap opera ‘Casualty’ describing this method in detail. The authors knew when this particular episode of Casualty was going to be broadcasted and interviewed suicide attempters in hospitals, before and after the airing of this episode, about their reason to commit suicide and whether or not the soap opera had influenced their decision. Hawton et al. (1999) found an increase in self-poisoning in the two weeks after the broadcast. Also increase in paracetamol overdoses was found, compared to non-paracetamol overdoses.
In Austria, Etzersdorfer, et al. (2004) found an increase in suicides by firearm in the weeks after a major Austrian newspaper reported a suicide by firearm by the owner of a well-known hotel in Vienna. Most other Austrian newspapers reported this event in a moderate way, but the largest national newspaper of Austria covered this event extensively and in a sensational way for a number of days, with explicit headlines and photographs included in the article. The increase in suicides was only found in regions in Austria with a high distribution rate of this particular newspaper. The authors called this a dose-response relationship in this study that they described as a natural experiment (Etzersdorfer et al., 2004).
Other authors (Phillips, 1974, and many of the authors who performed replicating studies) also reasoned that a contagion effect was not only found in the same week as the newspaper article but also in the weeks thereafter. In studies about the effect of advertising in the period following the publication of advert, this effect is called ‘decay’ (Joseph, 2006).
19
Hypothesis 3: News items (in newspapers and on television) about child suicide will influence the amount of conversations to the Kindertelefoon on the topic of suicide in the same week and the following weeks.
Hypothesis 3a: News items (in newspapers and on television) about child suicide will influence the amount of telephone conversations to the Kindertelefoon on the topic of suicide in the same week and the following weeks.
Hypothesis 3b: News items (in newspapers and on television) about child suicide will influence the amount of chat conversations to the Kindertelefoon on the topic of suicide in the same week and the following weeks.
At the moment, authors seem to agree there is enough evidence that demonstrates the existence of the Werther Effect and focus of research should be shifted to identifying certain characteristics of a story that are considered to lead to a copycat effect and raising awareness by the press on how to report responsibly on the issue of suicides (Gould, 2001, 2003; Pirkis & Blood, 2010; Sisask & Värnik, 2012; Stack, 2005). The next part of this chapter will elaborate on these characteristics.
2.1.1. Characteristics of news items leading to suicidal contagion
As mentioned in the previous paragraph, there is ample evidence on the existence of a copycat effect in suicides after press coverage of suicide acts. However, not all newspaper articles or television broadcasts on this subject lead to the same strength of the Werther effect. The strength of the effect depends on certain content characteristics of articles, which will be summarized briefly in this paragraph.
20 Research shows that imitative suicidal behaviour increases when coverage of suicide in the media is prominent, detailed, repeated, sensational, glorified, lengthy, and/or explicitly describes the method used (Gould, 2001; Hawton et al., 1999; Pirkis, Burgess, Blood, & Francis, 2007; Tousignant, Mishara, Caillaud, Fortin, & St-Laurent, 2005; Yip et al., 2006).
Examples of sensational articles may include photographs, big headlines (especially including the word ‘suicide’), are on the front page of a newspaper (Martin & Koo, 1997; Pirkis et al., 2007).
Furthermore, the Werther effect is particularly significant when a suicide by a celebrity is reported (Stack, 2003; Yip, et al., 2006), as opposed to news items about the suicide of a relatively unknown person, especially with entertainers who teenagers can relate to (Blood & Pirkis, 2001; Chen et al., 2010; Cheng et al., 2007; Yip et al., 2006).
After reviewing 42 studies, Stack (2000) showed that studies on the Werther effect in case celebrity suicides were 14.3 times more likely to find this effect than studies that did not include a celebrity suicide. Stack (2003) also found that the copycat effect mainly shows in the country where that celebrity is from and that stories about foreign celebrities have little or no impact in other countries. This may explain why Martin and Koo (1997) did not find an increase in suicides in Australia after the suicide of Kurt Cobain, the lead singer of the famous 1990s band Nirvana from the United States.
Another factor that determines the strength of the Werther effect after a celebrity suicide is the way in which the celebrity or the suicide act itself is described; when the celebrity is glorified, the copycat effect is bigger than when the celebrity or the act itself is disapproved of in the news item (Schmidtke & Häfner, 1988; Stack, 2003; Tousignant et al, 2005).
21 In 2010, Chen et al. investigated the influence of media reporting of the suicide of Ivy Li, a 24-year old, famous female Taiwanese entertainer on attempted suicides after the reporting started. The death of this entertainer was reported in a sensational and emotional way, including interviews with her parents and details on the method used. In the first three days after her death, Ivy Li’s suicide was extensively reported, followed by intermittent reports for another two weeks (Chen et al., 2010). Suicide attempters were interviewed by psychiatrists, and they were asked about any exposure to the reporting in the media of Ivy Li’s suicide and whether or not the attempted suicide by the participants was influenced by the media reports. Chen, et al. (2010) found evidence that when vulnerable people come in contact with reports by the media of a suicide by a celebrity, a copycat effect in suicides occurs. The authors found the risk of being influenced by the media was significantly higher in younger respondents and in men (Chen, et al., 2010).
In a meta-analysis, Gould (2001) reviewed available studies on the copycat effect of suicides and concluded that the strength of the Werther effect depends on the amount of publicity to a suicide case, whether or not the suicide method is described (in detail), and also the location of the newspaper article. She found that articles on the front page lead to a bigger copycat effect. In the studies that were reviewed, Gould (2001) also found an increase in suicides by the same method after publication of these details.
Based on literature and discussions with experts, O’Carroll and Potter (1994) of the Centers for Disease Control and Prevention (CDC), summarized the aspects of news coverage that can promote suicide contagion, in the following list:
presenting simplistic explanations for suicide,
22 providing sensational coverage of suicide,
reporting ‘how-to’ descriptions of suicide,
presenting suicide as a tool for accomplishing certain ends,
glorifying suicide or persons who commit suicide, and
focusing on the suicide completer’s positive characteristics’
The following hypotheses were derived from the theory and subsequent reasoning:
H4: News items (in newspapers and on television) about child suicide, that contain one of the contagion characteristics listed by O’Carroll and Potter (1994), will positively influence the amount of conversations to the Kindertelefoon on the topic of suicide.
o H4a: News items (in newspapers and on television) about child suicide, that contain
one of the contagion characteristics listed by O’Carroll and Potter (1994), will positively influence the amount of telephone conversations to the Kindertelefoon on the topic of suicide.
o H4b: News items (in newspapers and on television) about child suicide, that contain
one of the contagion characteristics listed by O’Carroll and Potter (1994), will positively influence the amount of chat conversations to the Kindertelefoon on the topic of suicide.
Pirkis et al. (2007) categorized the context of their collected reported news articles on suicides that give rise to suicidal contagion, into three groups:
‘suicides reported in a broader context’, such as
o systems failures, such as such as lack of appropriate funding, o stress situations at work,
23 o economic crisis,
o lack of support of prisoners in jail,
o insensitivity of bureaucracies such as child support agencies, and o economic downturn
‘suicides by celebrities’
o that have fallen from grace, or o that were glorified
‘suicides involving unusual circumstances/methods’, including o murder-suicides,
o suicides pacts
H5: News items (in newspapers and on television) about child suicide, that contain one of the context characteristics listed by Pirkis, et al. (2007), will positively influence the amount of conversations to the Kindertelefoon on the topic of suicide.
H5a: News items (in newspapers and on television) about child suicide that contain one of the context characteristics listed by Pirkis, et al. (2007), will positively influence the amount of telephone conversations to the Kindertelefoon on the topic of suicide.
H5b: News items (in newspapers and on television) about child suicide, that contain one of the characteristics listed by Pirkis, et al. (2007), will positively influence the amount of chat conversations to the Kindertelefoon on the topic of suicide.
In this paragraph, characteristics of new items that have a positive influence of the copycat effect in suicides were summarized. In the next paragraph, the preventive effect the media can also have, will be outlined.
24
2.2.
The Papageno effect: the preventive effect of the media on
suicides
As described above, a certain way of covering suicide stories by the media can lead to an increase in suicides (the Werther effect), but that does not mean that suicide cases should be banned from being reported by the media (Pirkis & Blood, 2010). In some cases it is appropriate to report a suicide, for example in order to educate people on the topics of suicide and depression. Nevertheless, an editorial staff must be beware of the risks of reporting on suicides and should balance the public’s ‘right to know’ against the risk of suicidal contagion, and it is advised that this topic is treated carefully (Pirkis & Blood, 2010). In Australia for example, only 1% of suicide cases were reported in the media over a one-year period. Stories that were considered newsworthy included mainly suicides by celebrities and suicides involving unusual or dramatic circumstances were reported (Pirkis, et al., 2007).
Lately, research has shifted from examining evidence on the Werther effect to researching the way in which the media can have a preventive effect on suicidal behavior. Niederkrotenthaler, et al. (2010) named the protective effect the media can have on suicidal behavior the ‘Papageno effect’. This effect will be described in this paragraph.
Papageno is a character in one of Mozart’s operas, who has suicidal thoughts since he feared to lose his loved one. In the end, he did not commit suicide because other people talked him out of it and showed him ways to cope with his fear (Niederkrotenthaler, et al. 2010).
One example of the protective effect that the media can have on suicidal behavior was given by Etzersdorfer & Sonneck (1998) in an unplanned field experiment. In the period following the introduction of the subway system in Vienna, Austria in 1978 there was a sharp rise in
25 suicides, especially by jumping in front of the subway. The media covered these events in an extensive way, with dramatic headlines and pictures of the persons that committed suicide (Etzersdorfer and Sonneck, 1998). After studying the literature on copycat suicides, the Austrian Association for Suicide Prevention started a press campaign in order to inform journalists on possible negative effects of articles on suicide, and to propose guidelines to cover suicides in a different way (Etzersdorfer & Sonneck, 1998). The authors found that media reporting did change after the campaign, with less dramatic articles and with some events not even covered. This has led to a 75% decrease in subway suicides (Etzersdorfer & Sonneck, 1998).
As mentioned earlier, in paragraph 1.1, Etzersdorfer, et al (2004) found that the suicide rate increased after reporting of a suicide case (by firearm) in Austria. At that moment in time, the guidelines were already in place, but one popular newspaper did not abide by these guidelines and reported on this case in a sensational way. This particular newspaper was distributed unevenly throughout the country. Etzersdorfer, et al. (2004) showed that 40% of the variance in suicides by gunshot could be attributed to the sensational way of reporting of this case by this newspaper.
Other authors have found a similar effect. In Switzerland for example, the amount of articles on suicide on the front page of newspapers decreased from 20% to 4%, and also the rate of sensational headlines decreased, from 62% to 25% leading, in turn, to a decrease in suicides. (Hawton & Williams, 2002; Michel, Wyss, Frey, & Valach, 2000; Wang, 2012; World Health Organization, 2014).
The World Health Organization (WHO), as well as the Center for Disease Control have introduced media guidelines on the reporting of suicide (O'Carroll & Potter, 1994; Sonneck,
26 Etzersdorfer, & Nagel-Kuess, 1994; Wang, 2012; World Health Organization, 2008). Also, in several other countries similar guidelines were introduced by government organizations or non-government organizations in order to convince the media to change the frequency and content of their suicide coverage in an effort to reduce copycat effects (Pirkis, et al., 2006; Chen, et al., 2010). However, most of these guidelines are not legally binding and research shows that journalist often are unaware of the existence of these guidelines and of the Werther effect in general. Authors therefore advocate collaboration between journalists, suicidologists and (non-) governmental organization when drafting guidelines. (Wang, 2012; Nepon, Fotti, Katz & Sareen, 2009; Pirkis, Blood, Beautrais, Burgess & Skehan, 2006; Pirkis, et al., 2007).
In The Netherlands, a media guideline was drafted and introduced by the Ivonne van de Ven Stichting in collaboration with journalists. The Ivonne van de Ven Stichting is a Dutch foundation that focuses on the prevention of suicide in The Netherlands. Their guideline is based on the recommendations of the World Health Organization amongst others. See appendix I for this guideline (in Dutch).
In general, the guidelines mentioned before encourage the media to decrease the prominence, detail, sensationalism of the reports on suicide and instead to describe suicide cases responsibly; in a concise and factual manner, including information for vulnerable individuals on where to turn to for help (Wang, 2012; O’Carroll & Potter, 1994; Pirkis, et al., 2007; Pirkis & Blood, 2010; World Health Organization, 2008, 2014).
27
2.2.1. Triggers for discussing suicidal thoughts
There are many possibilities for people to get help dealing with suicidal thoughts. Teenagers are said to be reserved about sharing personal information with adults, and even in situations when they are distressed, they rarely ask professionals for help (Boldero & Fallon, 1995; Offer & Schonert-Reichl, 1992). Confidentiality seems to be an important factor to teenagers when looking for help. Moreover, not all people that consider suicide as an option to deal with their problems are aware of their other options, including asking for professional help (Boldero & Fallon, 1995; Offer & Schonert-Reichl, 1992).
As mentioned previously, several media guidelines suggest that it is beneficial to include information on help-seeking options (for example telephone numbers of helplines) in news items about suicide.
By having follow-up conversations with people that have called a helpline, researchers showed that helplines are effective in reducing the risk of suicide among the callers (Gould, et al., 2007). Even though there is no conclusive evidence that suicide rates decrease because of the existence of helplines, helplines have proved to be useful (World Health Organization, 2014).
After Kurt Cobain, the front man of the famous 1990s band Nirvana committed suicide in April 1994, experts feared a copycat effect among teenagers. However, Cobain’s death had an opposite effect; less completed suicides than expected, but instead an increase in calls to helplines in the Seattle area where Cobain was from (Jobes, Berman, O’Carroll, Eastgard & Knickmeyer, 1996, Schonfeld, 2014a). The rise in calls may be caused by efforts of helplines
28 and by a change in media coverage after the introduction of media guidelines for the reporting of suicides, earlier that same month (Jobes et al, 1996).
A similar case is the recent suicide by actor Robin Williams, from the United States, in August 2014. His death was covered extensively by the media and led to an increase of calls to crisis hotlines in the United States (Szabo, 2014). One day after the suicide had happened, the National Suicide Prevention Lifeline received twice the amount of phone calls than on a regular day; around 7,400 calls instead of a typical 3,500 calls a day (Schonfeld, 2014b; Szabo, 2014). In the rest of the month of August 2014, there were 700-800 more calls than normal and also in the remainder of 2014 there were more calls than usual. On a smaller scale, local helplines also saw a rise in calls (Schonfeld, 2014b).
Sensational media coverage of suicides and especially suicides by celebrities can be harmful, as mentioned earlier. That is why there are many initiatives to urge journalists to cover these facts in a responsible manner, by focusing on the effectiveness of suicide prevention instead of the impact of the suicide itself. Responsible coverage, including information on where to go to for help, can raise awareness on the benefits of treatment and leading some people to get professional help instead of thinking about suicide as a solution (Schonfeld, 2014; Szabo, 2014).
Help-seeking in itself may also turn out to be contagious; when celebrities reveal in the media that they are looking for help, people may consider it an adequate alternative to actually committing suicide (Stack, 1987). Also, knowing someone who reached out to professional organizations for help may increase the likelihood that help-seeking is considered (Rickwood & Braithwaite, 1994).
29 Additionally, controversial topics tend to cause conversation. These topics are uncomfortable to discuss, but are on the other hand also more interesting, which increases the likelihood of discussion (Chen & Berger, 2013). The uncomfortable feeling when discussing a controversial topic like suicide, is suggested to be moderated by the anonymity of an internet forum or a helpline (Chen & Berger, 2013). According to Fukkink and Hermanns (2009) some children purposely choose the chat option to contact the helpline for children in the Netherlands (‘de Kindertelefoon’) instead of via telephone, as this medium makes it possible to discuss sensitive issues in private. In chat there is also more time to think about what you want to express and there is more time to interpret the answer. Especially emotional problems, including suicide or suicidal thoughts, are the subject matter of chat conversations, which is probably caused by so-called ‘channel reduction’; one cannot hear crying or stuttering via chat (Fukkink & Hermanns, 2009).
Hypothesis 6: News items (in newspapers and on television) about child suicide positively influence chat conversations to the Kindertelefoon about suicide to a larger extent than they influence telephone conversations to the Kindertelefoon on the same subject.
30
3. Research method
First a list of suicides by children publicized in the newspapers and via TV broadcasts was generated from the ‘Lexis Nexis Academic’ and ‘Beeld en Geluid’ databases respectively. See paragraph 3.1 and 3.2 for further details.
The detailed data on calls/chat conversations was collected from the Kindertelefoon. First this data had to be put in the right excel-format. After this, the data was imported into SPSS and all variables were recoded, as not all variables were entered in the same way in each year. A confidentiality agreement was signed for the use of the dataset of the Kindertelefoon and the dataset is not included in this document. However, some descriptives of the dataset are presented in paragraph 3.3.
Data from both sources were collated in Excel spreadsheets and analyzed using Excel (2013) and SPSS (version 22). The analysis examined the proportion of calls on the topic of suicide compared to all topics to the Kindertelefoon, and also a more detailed examination of the isolated topic of suicide.
The independent variable is de number of news items (initially both TV and newspapers combined, and later separately analyzed) about child suicide, per week. The dependent variable is the amount of conversations (telephone and chat combined or separated) to the Kindertelefoon on suicide in this case.
3.1.
Selection of newspaper articles on child suicide
Via the Dutch Lexis Nexis Academic-database (http://academic.lexisnexis.nl/) a first selection was made with the keywords ‘girl’ AND ‘suicide’, OR ‘boy’ AND ‘suicide’ (in Dutch: ‘meisje’ en
31 ‘zelfmoord’ or ‘jongen’ en ‘zelfmoord’) within a sentence. Those keywords in the same paragraph instead of the same sentence led to too many irrelevant articles (for example an item with the translated title “Father committed suicide. He left a family behind, including a 14-year old girl”). The newspaper articles can come from both Dutch national or regional newspaper within the given timeframe.
The timeframe of the articles included was from January 1994 up until September 2013, as this was also the timeframe of the available data from the Kindertelefoon.
After the initial selection, the articles were read and checked for relevance. In order to be relevant for this research, it had to mention an unmistakable suicide. Excluded were articles on non-fatal suicide attempts and dealing with suicide in general (for example an article about a policy program or an article about suicide rates in general), as opposed to on a specific case. The mentioning of suicide can be in the headline as well as in the body of the text. Not only front page articles were taken into account, also articles from other parts in the newspaper were relevant. Next to that, the age of the person(s) that had committed suicide had to be mentioned and it had to be under eighteen years. Only suicides by children were included, as they are said to be more susceptible to media influences and suicidal contagion (Baume, et al., 1997; Chen, et al. 2010; Gould, et al., 2003; Stack, 2003), and as they are the target group of the Kindertelefoon.
The body of the text was read and if applicable other details were recorded. Not all details are relevant for the current research. Some were added only to be able to trace the original newspaper article if needed and/or for additional research. The recorded details are:
32 Date of publication
Name of newspaper
National or Regional newspaper
The coverage of the newspaper
Title of the newspaper article
Gender of the person that committed suicide (if mentioned)
Age of the person that committed suicide
Name of the person that committed suicide (if mentioned)
Date of the suicide
Geographic location of the suicide (if mentioned); The Netherlands of another country.
Reason of the suicide (if mentioned). This varied for example from: o Bullying (at school and/or via social media)
o Thrill to kill a lot of people and the feeling that violence is fun (applicable for murder-suicides mostly) and/or the desire to become famous.
o Financial problems
o Shame following sexual violence (e.g. after a rape, some girls felt shame and their changes to get married declined. They could not handle the shame anymore and ended their lives)
o Love (for example after a break-up, or when a couple is not allowed to be together) o Fights with parents
o Gaming addiction (some children were reported to be unable to understand the difference between games and real life and thought they could live on in cyberspace (article in the Volkskrant of October 20, 2005). Also, in September 2007
33 some articles reported a suicide by a boy whose gaming account was hacked so he could not play anymore).
o Sympathy with terrorist organization. Several articles mentioned a boy who sympathized with Al Qaeda, and decided to kill people before killing himself. Also an article mentioned a murder-suicide on Hitler’s birthday.
o Pressure from parents/school over performance which some children could not handle or live up to anymore
o Psychological distress/psychiatric problems (including parents getting divorced and psychiatric illness and depression)
Method (if mentioned). This includes: o Hanging
o Shooting
o Drug overdose/poison intake (e.g. pesticides, acid)
Also the context of the article was categorized (if applicable) using the categories defined by Pirkis (2007). Pirkis found that only 1% of suicides in Australia was reported over a certain period and examined what type of suicide were considered newsworthy and found that there is evidence that certain cases were over-reported; for example suicides involving dramatic methods. Pirkis (2007) categorized the articles in three non-mutually exclusive groups:
Suicides reported in a broader context, such as:
o Systems failures, such as such as lack of appropriate funding, o Stress situations at work (mostly not applicable to children) o Economic crisis,
34 o Insensitivity of bureaucracies such as child support agencies.
o Economic downturn
Suicide by a celebrity
Suicides involving unusual circumstances/methods, including o Murder-suicides,
o Suicide pacts
The articles were also categorized (if applicable) in the contagion factors of the Center of Disease and Control (CDC, 1994). The Center of Disease and Control believes that the likelihood of suicide contagion may be increased by the following factors:
Presenting simplistic explanations for suicide. Most persons who have committed suicide have had a history of problems and suicide is never the result of a single factor or event.
Repetitive reporting of suicide. Excessive reporting tends to promote and maintain a preoccupation with suicide among at-risk persons, especially among persons 15-24 years of age.
Sensational coverage of suicide. This tends to increase the public’s preoccupation with suicide. It includes the mentioning of morbid details and publication of dramatic photographs (e.g. photographs of the funeral, the deceased person's bedroom, and the site of the suicide).
Reporting "how-to" descriptions of suicide. This includes technical details of the method. For example, reporting that a person died from carbon monoxide poisoning may not be harmful; however, providing details of the mechanism and procedures used to complete the suicide may facilitate imitation of the suicidal behavior by other at-risk persons.
35 Presenting suicide as a tool for accomplishing certain ends. Presentation of suicide as a means of coping with personal problems (e.g., the break-up of a relationship) may suggest suicide as a solution to at-risk persons.
Glorifying suicide or persons who commit suicide. News coverage is less likely to contribute to suicide contagion when reports of community expressions of grief are minimized.
Focusing on the suicide completer's positive characteristics. Empathy for family and friends often leads to a focus on reporting the positive aspects of a suicide completer's life. For example, friends or teachers may be quoted as saying the deceased person "was a great kid" or "had a bright future," and they avoid mentioning the troubles and problems that the deceased person experienced. Suicide may appear attractive to at-risk persons, especially when they rarely receive positive feedback.
3.1.1. Descriptives of the dataset: Newspaper articles
The articles from the initial search in the Lexis Nexis Academic database, were scanned for relevance. Ultimately, 180 newspaper articles about child suicide were included in the present research (see Appendix II, table 1). For a full list of the news items (TV and newspapers combined, see Appendix I.
Of the 180 newspaper articles, the same amount of articles were about suicides by boys as about suicides by girls. A small percentage included a group of children who committed suicide together (see Appendix II, table 2).
36 Most articles were published in regional newspapers (53.3%, see Appendix II, table 3). Some suicide cases led to articles in several different newspapers.
Not all newspapers have the same coverage or distribution rates. A national newspaper has a wider coverage than a regional newspaper. The coverage, measured as the number of copies distributed per year was added to the dataset (see Appendix II, table 4). Additionally, a fixed coverage rate (the average coverage per newspaper title) is included (see Appendix II, table 5). There were not many TV broadcasts that fit the criteria to be included in the dataset. But reasoned in the same period a newspaper article was published about a child suicide was also attention in other media about the same case. The more attention that was given to a case in newspapers, also more attention in other media is expected. Consequently, the coverage of the newspapers serves as an index for the size/range of the newspaper, and the amount of people that heard in one way or another about a suicide case; when a suicide case is discussed in the surroundings of the child, the child comes in contact with the subject, which may trigger him/her to contact the Kindertelefoon.
It is expected that a newspaper article does not only cause an effect in the same week, but also in the subsequent weeks. A decay rate of 50% of the coverage figures is taken into account per week.
The majority of cases (57.2%, see Appendix II, table 6) concerned child suicides by children from other countries than the Netherlands. Most newspaper articles were on children aged 13 to 15 (55.6%, see Appendix II, table 7). Bullying is the most frequently mentioned reason of child suicide in the newspaper articles (48.3%, see Appendix II, table 8). Note that the reason of the suicide is not always known/mentioned (in 33 cases, see Appendix II, table 8).
37 Also, the context categories derived from Pirkis et al. (2007) show bullying as the most frequently mentioned context category (39.1% see Appendix II, table 9). Note that in 18.3% of the cases a context category was not applicable (according to the classification by the author). For a full list of the newspaper articles, including the context classification, see Appendix I.
The classification into the contagion categories (derived from Carroll & Potter, 1994), show that most articles are of a sensational nature or present the suicide as a tool for accomplishing a certain end (i.e. problem solving), see Appendix II, table 10. Note that 44% of the articles did not have a contagious content (according to the classification by the author, based on the categories as described by Carroll & Potter, 1994). For a full list of the newspaper articles, including the contagion classification, see Appendix I.
3.2.
Selection of television broadcasts on child suicide
Next to newspaper articles, also television broadcasts on suicides by children under the age of nineteen were taken into account. Expected was that children are more likely to watch TV than to read newspapers. The broadcast were selected by searching in the archive database of ‘Beeld en Geluid’ (www.beeldengeluid.nl). This database contains information on broadcasts by the public TV channels. Unfortunately, this database does not include the news-broadcasts, but only programs such as current affairs, documentaries and discussion shows. As this database includes only the public channels, there is no data on private broadcasting. Consequently, the available data on TV broadcasts on child suicide is limited.
38 Via ‘Beeld en Geluid’ a first selection was made with the keywords ‘suicide’ AND ‘youth’ (in Dutch: ‘zelfmoord’ en ‘jeugd’). These keywords were predefined by the database. The same timeframe as the selection of newspapers was used, i.e. from January 1994 up until September 2013.
After the initial selection, the broadcasts were checked for relevance by reading the description. In order to be relevant for this research, it had to mention an unmistakable suicide. Non-fatal suicide attempts were left out of the selection. Next to that, the age had to be mentioned and it had to be under eighteen years. The description was read and other details were recorded. These details include:
Date of broadcast
Name of network
Type of TV program show (for example: documentary or current affairs program)
Title of the broadcast
The other details that were recorded are the same as the details of the newspaper articles, see paragraph 3.1.
3.2.1. Descriptives of the dataset: TV broadcasts
During the period January 1994 to September 2013 there were only six broadcasts on child suicide that fit the criteria. All six broadcasts concerned suicides by boys (see Appendix II, table 11). Four broadcasts were on child suicide in The Netherlands, while two were on suicides by children outside The Netherlands (see Appendix II, table 12). All cases were on children aged 14 and over (see Appendix II, table 13). The categories of Context (derived from Pirkis et al., 2007) show only general categories (see Appendix II, table 14). The classification
39 into the contagion categories (derived from Carroll & Potter, 1994), show that most broadcasts are of a sensational nature or emphasize the positive characteristics of the deceased. For a full list of all TV broadcasts about child suicide, including the context and contagion categories, see Appendix I.
3.3.
Database Kindertelefoon
3.3.1. Background on the Kindertelefoon
The history of the Kindertelefoon will be outlined briefly in this paragraph.
In 1979, which was declared the International Year of the Child by UNESCO, a group of parents founded a children’s helpline in The Netherlands, and called it ‘De Kindertelefoon’. The Kindertelefoon targets children aged eight to eighteen (Kindertelefoon, 2013).
The reason of being of the Kindertelefoon, from the start and still up to date, is the right of every child to express their opinion and grow up to be a healthy adult. The Kindertelefoon always respects the child and it offers children easy accessible help, emotional aid, information and advice. Next to that, the Kindertelefoon observes trends in individual questions and addresses those issues at the appropriate bodies (Kindertelefoon, 2013).
The Kindertelefoon started out relatively small; receiving phone calls at only one location, and solely manned by volunteers. The organization grew to 18 locations throughout The Netherlands in 2014, with approximately 800 volunteers. As of January 2015, all regional offices of the Kindertelefoon have been united in a national foundation; ‘Landelijke Stichting De Kindertelefoon’ (Kindertelefoon, 2013).
40 During the early years, the only possibility to contact the Kindertelefoon was by landline. Calls from landlines became free of charge from 1992, which caused a remarkable increase in calls (Kindertelefoon, 2004). A lot has changed during the past decades in the area of communication; cellular phones became more common, for adults as well as for children, and children’s internet is very common (Statistics Netherlands, 2010). In 1995 the website, www.kindertelefoon.nl, was launched, and was updated in 2012. As from 2002 calls from mobile phones to the Kindertelefoon became free of charge. In 2003, a chat service was introduced which was an immediate success, even without many promotional activities (Kindertelefoon, 2003). As from 2008 it was made possible to immediately put phone calls through to other assistance bodies in emergency situations, for instance in a case of child abuse (Kindertelefoon, 2012).
All questions are taken seriously and volunteers always stimulate callers/chatters to choose the best fitting solution themselves, sometimes referring them to other assistance bodies. There is not necessarily a solution to every problem; sometimes children only call for the sake of having a conversation, as talking to someone is sometimes enough help in itself (Kindertelefoon, 2012).
3.3.2. Topics of conversations
A research, conducted in 2004 by ‘Kaboem Media’, showed that 95% (n = 2210) of the interviewed children between the ages of 9 and 16 is familiar with the name ‘Kindertelefoon’. Over 50% of the respondents, especially the between the ages of 9 and 13, indicated to consider calling the Kindertelefoon in case of trouble. 18% of the respondents said that they have called the Kindertelefoon once or more in the past (Kindertelefoon, 2004).
41 All calls and chat conversations are registered in a centralized database. In 2013, there were 116,604 telephone conversations and 71,712 chat conversations (Kindertelefoon, 2013). Of the children who contacted the Kindertelefoon in 2013, either by phone or by chat, the majority was aged 13-15. Also can be noticed that both callers and chatters are mostly female (23% in 2012). There are also groups of children who initiate calls or chat conversations, but proportionally this group is much smaller than single callers (4.5% in 2012). Next to that there is a group of callers/chatters whose gender and/or age is unknown (2% in 2012) as children are not required to provide this information (Kindertelefoon, 2012). Even though the Kindertelefoon kept record of all conversations, in terms of age, gender, duration of conversation and topic of conversation, all conversations are treated anonymously. No record is kept of the name, cultural background, address and/or phone-number of the caller. For this reason it is not possible to differentiate one-time users from children who have used the services of the Kindertelefoon more than once. Requests from children that are calling or chatting for the second time, wanting to talk to the same volunteer cannot be granted.
During the early years, when there were only telephone conversations, the most common topics were home/family issues, sexuality and love/dating (Kindertelefoon, 2005). Over the years, the most discussed topics did not change much, but there is a difference in which topics are mostly discussed via telephone and which topics are mostly discussed via chat (see Appendix II, table 16).
For most conversations a ‘Main topic’ was registered, as well as a ‘Subtopic’, which is a more detailed specification of the main topic (see Appendix II, table 17 for an overview of all main topics and the corresponding subtopics).The main topic was not filled out in every occasion. In the cases where it was filled out, in most years, ‘G. Sexuality’ is the most talked about
42 subject of conversation, for girls, boys as well as groups, followed by ‘B. Relationships’ (see Appendix II, chart/graph 1 and Table 18 all years combined).
Suicide (subtopic C02) is categorized by the Kindertelefoon under the overall topic of ‘Emotional Issues’, and was the main topic of conversation 397 times via chat in 2013 and 131 via telephone in 2013 (database Kindertelefoon). It is not the most common subject, but as also mentioned in the first section of this paper, there is an increase in suicides amongst teenagers, and the Kindertelefoon’s volunteers find the subject of suicide one of the most difficult to deal with (Kindertelefoon, 2007). Main topic ‘C. Emotional issues’, which includes suicide was registered 55,320 times. This amounts to 2.6% of all valid cases (see Appendix II, table 18).
Topic ‘M. Other’ is category that included all topics that did not fit in any of the topics from A to L. It is excluded from this research.
3.3.2.1. Subtopics
In 971,980 cases a subtopic was registered. This amounts to 37.4% of total (serious) conversations. Of the registered topics, 0.8% of the conversations was about suicide. It is the 38th most used sub topic. It amounts to a total of 8,020 conversations (see Appendix II, table 19). There are no distinct changes in the share of subtopic suicide on the total amount of registered subtopics over the years (see Appendix II, table 20 and chart/graph 2).
