Is the narrative the message? The relationship between suicide-related narratives in media reports and subsequent suicides

Objectives: When journalists report on the details of a suicide, the way that they contextualize the meaning of the event (i.e. the ‘narrative’) can have significant consequences for readers. The ‘Werther’ and ‘Papageno’ narrative effects refer to increases and decreases in suicides across populations following media reports on suicidal acts or mastery of crises, respectively. The goal of this study was to investigate the impact of these different narrative constructs on subsequent suicides. Methods: This study examined the change in suicide counts over time in Toronto, Canada. It used latent difference score analysis, examining suicide-related print media reports in the Toronto media market (2011–2014). Articles (N = 6367) were coded as having a potentially harmful narrative if they described suicide in a celebrity or described a suicide death in a non-celebrity and included the suicide method. Articles were coded as having potentially protective narratives if they included at least one element of protective content (e.g. alternatives to suicide) without including any information about suicidal behaviour (i.e. suicide attempts or death). Results: Latent difference score longitudinal multigroup analyses identified a dose–response relationship in which the trajectory of suicides following harmful ‘Werther’ narrative reports increased over time, while protective ‘Papageno’ narrative reports declined. The latent difference score model demonstrated significant goodness of fit and parameter estimates, with each group demonstrating different trajectories of change in reported suicides over time: (χ2[6], N = 6367) = 13.16; χ2/df = 2.19; Akaike information criterion = 97.16, comparative fit index = 0.96, root mean square error of approximation = 0.03. Conclusion: Our findings support the notion that the ‘narrative’ matters when reporting on suicide. Specifically, ‘Werther’ narratives of suicides in celebrities and suicides in non-celebrities where the methods were described were associated with more subsequent suicides while ‘Papageno’ narratives of survival and crisis mastery without depictions of suicidal behaviours were associated with fewer subsequent suicides. These results may inform efforts to prevent imitation suicides.

the latent difference score, ΔSuicide(t) n . If observations occur at fixed intervals, the time between pairs of latent scores is constant (i.e., Δt = 1) and the latent difference score can be interpreted as the rate of change of the true score (ΔSuicide(t) n /Δt = ΔSuicide(t) n .)

Univariate LDS Models
Using the latent rate of change (ΔSuicide(t) n /Δt) as the outcome variable, there are several ways to model univariate longitudinal change (Hamagami and McArdle, 2001;McArdle and Nesselroade, 2003). In the "dual change" model, suicide change over time can be expressed as: ΔSuicide(t) n = α Suicide x s Suicide,n + β Suicide x Suicide (t -1) n .The additive change component (α Suicide x s Suicide,n ) involves change which is constantly related to a score s Suicide,n . This is a latent variable, involving values that vary across subjects but are constant over time. The coefficient α Suicide can be considered as a factor loading, and is usually fixed to 1 for identification purposes. The s Suicide,n term is the intercept term which can change from subject to subject. The proportional change component (β Suicide x Suicide (t -1) n ) involves change that is proportional to the previous latent score. β Suicide indicates the proportional effect of a previous latent variable on the subsequent rate of change, and can be either time-invariant, or time-varying (i.e., β Suicide (t)).
Simplifying the dual change model leads to three models of univariate change. In the constant change score model, the β Suicide coefficient is set to zero, and latent change is constant within a subject over time.
The resulting equation would be: ΔSuicide(t) n = α x s Suicide,n , β Suicide = 0 In the proportional change score model, the  Suicide coefficient is set to zero, and latent change is proportional to the latent score from the previous time point.

Australian and New Zealand Journal of Psychiatry
In the no change score model, the latent scores do not change over time. However, the observed scores may vary over time due to the random error term, e(t) n . The resulting equation would be: ΔSuicide(t) n = 0, α Suicide = β Suicide = 0 Figures 1a, 1b, 1c and 1d represent the four univariate LDS models as path diagrams used in structural equation modelling. In these path diagrams, a latent variable (represented as a circle) is comprised of an observed variable (represented as a square), with associated measurement error (represented as a circle). A directional arrow symbolizes the influence of one latent variable on another, while a bi-directional arrow represents correlations.
1. Note. Suicidality = Number of suicide deaths reported. 0 (=) indicates parameter is not estimated. "p close fit" = p value for testing the null hypothesis that the population RMSEA is not greater than .05; CFI = comparative fit index; AIC = Akaike information criterion. E(sn) = additive change coefficient; β = proportional change coefficient. The α and β coefficients resulting equation characterizes the change in Suicide using two components: additive change (i.e., α suicide x s suicide,n ), and proportional change (i.e., β suicide x suicide (t -1). Based on the modelling results, an equation can be generated that describes the longitudinal model based on each of the variables: ΔSuicide(t) n = α suicide x E[s suicide , n ] + β s x E[Suicide (t -1) n ] 2. a p < .05. b p < .01. c p < .001.