Selected Research Papers in Social Change, Education, Labour Market, and Criminology Volume II

Table 7: Multivariate analysis of variance regarding the consequences of class membership (n = 311; calculated in the framework of the general linear model)

Polytropicintensive stalking Polytropicmoderate stalking Distant stalking Surveillance

External stalking consequences mean number 3.46 2.12 0.88 1.05

Univariate test: F = 37.61; p = .000; partial η ² = .27 Internal stalking consequences mean number 6.09 4.50 2.96 2.95

Univariate test: F = 27.51; p = .000; partial η ² = .21 Global multivariate test Pillais test = 0.30; F = 17.77; p = .000

All things considered, the findings support the hypothesis that with a rising versatility and intensity of the behavioral contents of stalking, the probability of the victim being left with tangible damages increases continuously. Parallel to this, they substantiate confidence in the findings of the LCA and the characterization of the clusters. In terms of a construct validation, one may infer from the confirmation of the hypothesis that with the four class structure identified here, we were able to develop a stalking typology which is commensurate with reality, substantially convincing, and empirically useful.

4 Summary

The focus of the present article was on the segmentation of a population of stalking victims founded upon the phenomenology of their victimization. On the basis of survey data from 311 victims of obsessive pursuit and harassment living in Upper Austria, an investigation was conducted to determine whether one can condense the mass of different manifestations of stalking into clearly delineated victimization profiles. With the assistance of a latent class analysis, four distinct victimization patterns could be identified:

• a polytropic-intensive form of stalking, which is distinguished by a very broad range of stalking-related offensive behaviors and where, in many cases, the offender does not even refrain from violent action against the victim and/or his/her private property,

20 The pairwise contrasts were tested for significance with Scheffé tests. According to the results of these tests, the victims of distant stalking and surveillance victims can not be significantly differentiated from each other with regard to either the external (p = .983) or internal (p = .999) effects of stalking.

Patterns of Stalking Victimization: A Behavioral Typology 128

• a polytropic-moderate type of stalking, for which the infractions also prove to be relatively diverse, but which generally does not involve invasions of the victim’s privacy or physical assault,

• a distant stalking, which is dominated by persistent attempts to establish contact with the victim through various means of communication, and finally

• stalking as pure surveillance, whereby the offender focuses on gathering information about the victim. With a view to their prevalence, the polytropic-moderate phenotype of stalking appears to be the most widespread among those stalking cases which have been brought to the attention of the police: About four out of every ten reported stalking incidents can be placed in this category. Three out of ten reports pertain to polytropic-intensive stalking; two out of ten charges result from distant stalking. Stalking which is solely limited to surveillance activities is seldom reported to the police: Only 6 % of the observed victimization profiles corresponded to this type.

As expected, the stalking profile has a substantial impact on the victim’s wellbeing. The number of disadvantages suffered by the victims increases steadily from surveillance and distant stalking – these two types are comparable in terms of the damages incurred on the victim – via polytropic-moderate up to polytropicintensive stalking. Such a relationship can be demonstrated for both the external (physical, social and economic) and the internal (psychological) consequences of stalking.

The behavioral profile characterizing an incident of stalking is itself strongly influenced by the type of prior relationship between offender and victim. Stalking by a former sexual partner harbors the greatest potential for escalation. The reasoning behind this may well be that the inhibitions which prevent persons from invading the private and intimate spaces of others are less pronounced when it comes to former sexual partners in comparison to less familiar persons. Another explanation may be that ex-partner-stalking is often the continuation of a previously existing violent and abusive relationship. Table 8 reiterates, on a descriptive level, the connection between victim-offender relationship and victimization profile. It becomes immediately apparent that obsessive pursuit by a former sexual partner often assumes the form of polytropic-intensive stalking, which entails the most severe impairment of the victim’s quality of life.

Table 8: Stalking profile by offender–victim relationship (n = 266)

Victimization profile Former sexual partner Other stalker

Polytropic-intensive stalking 56 % 23 % Polytropic-moderate stalking 35 % 41 % Distant stalking 7 % 27 % Surveillance 2 % 9 % Total 100 % 100 %

Cramer’s V = .38; p = .001

Stalking prevention that is motivated primarily out of concern for the welfare of the victims should place particular emphasis on incidents of obsessive pursuit and harassment that emerge from prior romantic relationships. These types of

Patterns of Stalking Victimization: A Behavioral Typology 129

incidents, which are generally accompanied by a broad array of infringements including serious crimes, wreak the most damage among their victims. With a demonstrated proportion of 38 % of all of the stalking cases reported to the police, they represent an impressive potential for preventative measures which, at least in Austria, is still waiting to be capitalized on.

References
Akaike, H. (1987): Factor Analysis and AIC. Psychometrika 52: 317–332. Bacher, J. (1996): Clusteranalyse. München: Oldenbourg. Bortz, J./Döring, N. (1995): Forschungsmethoden und Evaluation. Berlin: Springer. Budd, T./Mattinson, J. (2000): The Extent and Nature of Stalking: Findings from the 1998 British Crime Survey. London: Home Office. Bühl, A. (2008): SPSS 16. Einführung in die moderne Datenanalyse. München: Pearson. Cupach, W./Spitzberg, B. (2004): The Dark Side of Relationship Pursuit: From Attraction

to Obsession and Stalking. Mahwah: LEA. De Fazio, L. (2009): The Legal Situation on Stalking Among the European Member States. European Journal on Criminal Policy and Research 15: 229–242. Dressing, H./Kuehner, C./Gass, P. (2006): Stalking in Deutschland. In: Hoffmann, J./Voß,

H.G. (eds.): Psychologie des Stalking. Grundlagen – Forschung – Anwendung. Frankfurt: Verlag für Polizeiwissenschaft. 27–43.

Hagenaars, J./McCutcheon, A. (2002): Applied Latent Class Analysis. Cambridge: Cambridge University Press. Hoffmann, J. (2006): Stalking. Heidelberg: Springer. Johnson, M./Kercher, G. (2009): Identifying Predictors of Negative Psychological

Reactions to Stalking Victimization. Journal of Interpersonal Violence 5: 866–882. Lo, Y./Mendell, N./Rubin, D. (2001): Testing the Number of Components in a Normal

Mixture. Biometrika 88: 767–778. Löhr, G. (2008): Zur Notwendigkeit eines spezifischen Anti-Stalking-Straftatbestandes in

Deutschland. Eine rechtsvergleichende Untersuchung unter besonderer Berücksichtigung der australischen Rechtsentwicklung. Berlin: Duncker & Humblot.

Magidson, J./Vermunt, J. (2004): Latent Class Models. In: Kaplan, D. (ed.): The Sage Handbook of Quantitative Methodology for the Social Sciences. London: Sage. 175–198.

McCutcheon, A. (1987): Latent Class Analysis. London: Sage. McLachlan, G./Peel, D. (2000): Finite Mixture Models. New York: Wiley. Meloy, J. (1996): Stalking (Obsessional Following): A Review of Some Preliminary

Studies. Aggression and Violent Behavior 2: 147–162. Mitgutsch, I. (2009): Aspects of “Stalking” in Austrian Criminal Law. In: Bacher, J./Gorniak,

J./Niezgoda, M. (eds.): Selected Research Papers in Education, Labour Market and Criminology. Volume 1. Linz: Trauner. 110–122.

Mullen, P./Pathé, M./Purcell, R. (2000): Stalkers and Their Victims. Cambridge: University Press.

Mullen, P./Pathé, M./Purcell, R./Stewart G. (1999): Study of Stalkers. American Journal of Psychiatry 156: 1244–1249.

Muthén, L./Muthén, B. (1998–2007): Mplus User’s Guide. Los Angeles: Muthén & Muthén. Orme, J./Combs-Orme, T. (2009): Multiple Regression with Discrete Dependent

Variables. Oxford: University Press. Pathè, M./Mullen, P. (1997): The Impact of Stalkers on their Victims. British Journal of Psychiatry 170: 12–17.

Patterns of Stalking Victimization: A Behavioral Typology 130

Schwartz, G. (1978): Estimating the Dimension of a Model. The Annals of Statistics 6: 461–464.

Sclove, L. (1987): Application of Model-Selection Criteria to Some Problems in Multivariate Analysis. Psychometrika 52, 333–343.

Sheridan, L./Blaauw, E. (2006): Stalkertypologien und Interventionsstrategien. In: Hoffmann, J./Voß, H. (eds.): Psychologie des Stalking. Grundlagen – Forschung – Anwendung. Frankfurt: Verlag für Polizeiwissenschaft. 73–91. Sheridan, L./Boon, J. (2002): Stalker Typologies: Implications for Law Enforcement. In: Boon, J./Sheridan, L. (eds.): Stalking and Psychosexual Obsession. Chichester: Wiley. 63–82.

Spitzberg, B. (2002): The Tactical Topography of Stalking Victimization and Management. Trauma, Violence, and Abuse 3: 261–288. Spitzberg, B./Cupach, W. (2003): What mad pursuit? Obsessive Relational Intrusion und Stalking Related Phenomena. Aggression and Violent Behavior 8: 345–375.

Spitzberg, B./Cupach, W. (2007): The State of the Art of Stalking: Taking Stock of the Emerging Literature. Aggression and Violent Behavior 12: 64–86.

Tjaden, P. (2009): Stalking Policies and Research in the United States: A Twenty Year Retrospective. European Journal on Criminal Policy and Research 15, 261–278. Tjaden, P./Thoennes, N. (1998): Stalking in America: Findings From the National Violence Against Women Survey. Washington: National Institute of Justice.

Turner, J./Thayer, J. (2001): Introduction to Analysis of Variance: Design, Analysis, and Interpretation. London: Sage.

Voß, H./Hoffmann, J./Wondrak, I. (2006): Stalking in Deutschland. Aus Sicht der Betroffenen und Verfolger. Baden-Baden: Nomos.

Wondrak, I./Meinhardt, B./Hoffmann, J./Voß, H. (2006): Opfer von Stalking – Ergebnisse der Darmstädter Stalkingstudie. In: Hoffmann, J./Voß, H. (eds.): Psychologie des Stalking. Grundlagen – Forschung – Anwendung. Frankfurt: Verlag für Polizeiwissenschaft. 45–72. Zona, M./Sharma, K./Lane, J. (1993): A Comparative Study of Erotomanic and Obsessional Subjects in a Forensic Sample. Journal of Forensic Sciences 38: 884–903.

Patterns of Stalking Victimization: A Behavioral Typology 131

David Kemethofer
Criminal and Antisocial Behaviour in Upper Austrian Print Media

Results of a Content Analysis

1 Introduction

Criminal and antisocial behaviour is part of society and empirically measurable. There are a variety of approaches available for this purpose. The most common method used to analyse criminal and antisocial behaviour is the traditional crime rate. However, the result of such analyses is a distorted reality due to unreported or improper claims. Dark field research methods have been developed to minimise this problem. According to Kania (2004), the subjective feeling is another way of measuring criminal and antisocial behaviour. This kind of individual apperception is influenced by media reporting (see Kania, 2004: 137f). Wilms states that consuming media leads to certain knowledge about the appearance of crime which does not reflect real life. (see Willms, 2004: 523ff)

Some earlier studies (Derwein, 1996; Scharf et al., 1999) already observed that the reporting of criminal and antisocial behaviour in newspapers is distorted. An inaccurately or inappropriately presented picture in the media may increase the general fear of criminal and antisocial acts as well as support what are arguably the wrong political measures to defeat crime. (see Schneider, 1980: 211f)

As the studies mentioned were based on the analysis of German newspapers it is sociologically and criminologically important to explore the content of Upper Austrian newspapers, too. Here it is essential to have a look at the image portrayed of offenders and victims as well as the way criminal and antisocial behaviour is described. To establish in which way public opinion could be influenced concerning the reasons for crime and political measures, this study also explores causes of deviant behaviour and measures offered to defeat crime.

For the purposes of this study, content analysis is the best way to address these issues. A codebook with 260 variables was compiled and a sample featuring more than 1,500 articles ranging from 2004 to 2007 was collected. It was important for every article chosen to be related in some way to Upper Austria and also have the possibility to influence the subjective feeling of the readers.

The following pages present an overview of the background of the study. Firstly we will have a look at the media system in Upper Austria with a focus on the chosen newspapers. As we assume that the impact of media is important for this study, the history of media effect-research is likewise discussed. Following on from this, the reader will find the main features of the content analysis and the results and conclusions drawn from this study.

Criminal and Antisocial Behaviour in Upper Austrian Print Media 132

2 Newspapers in the Upper Austrian media landscape

Newspapers have existed for centuries. Their role as a mass medium is – in spite thof the introduction of new types of media such as television and the internet – still an important one. Compared with different types of mass media the tradition of newspapers is quite old and started in the early 17th century when the first periodical newspapers appeared in Europe. Modern newspapers evolved during the 181 century. There are now many different types of newspapers, some with a very local flavour and others which are free of charge or published just once a week. In this context print media is very important in Upper Austria, too. In this study seven different newspapers with a strong connection to the region were chosen for the analysis. To be more specific, the collection includes three daily newspapers, two weekly newspapers23 which every household receives automatically and two free newspapers which are available at railway stations or on the tram. Every newspaper had a local editorial office to guarantee the local context of the articles. Furthermore, the newspapers had to have a certain minimum circulation. As a result, newspapers of a higher quality were eliminated from the process. (see Giddens, 1999: 397ff; Wilke, 2002: 460ff)

In Upper Austria all newspapers together reached a total circulation figure of 71.6% in 2007, whereby the Sonntagsrundschau reached the highest level with 60%. One reason could be the free issue on Sundays. On the whole people have more free time on Sundays and therefore more time to read newspapers. Furthermore, the coverage of both free newspapers was lower than 10%, whereas every daily newspaper reached at least one-quarter of society, with a peak value of 43.9% recorded for Neue Kronen Zeitung, a value which cannot be found in any other European country and warrants further discussion regarding the power of a single newspaper. (see Fritz/Klingler, 2006: 223ff; Günther/Hüffel, 1999: 13; Seethaler/Melischek, 2006: 337ff)

3 Media impact

Besides the circulation of a certain newspaper it is also important to look at the thpossible effects of media or, in other words, how reporting in a certain way can influence public opinion. In sociology, interest in the effects of mass media started in the early 20 century. Max Weber (1997 [1911]) already asked to what extent newspapers are able to contribute to cultural goods and how modern people are influenced by them. (see Weber, 1997: 138ff)

Initial simple stimulus-response models, influenced by other sciences, have been used to explain possible effects. In the meantime, this approach has been confuted and agenda-setting models are state-of-the-art in explaining the ways

1 Neue Kronen Zeitung, Oberösterreichische Nachrichten, Österreich

2 Linzer Rundschau, Sonntagsrundschau

3 Heute, Oberösterreichs Neue

Criminal and Antisocial Behaviour in Upper Austrian Print Media 133

mass media influences public opinion. This approach understands the possibility of media effects in a long-term way, beginning with the subjective selection of topics reported by the newspapers. This selection is based on the news value. The news value determines the likelihood of publication. (see Jäckel, 2008: 169ff; Schulz, 2002: 355f)

In a further step this means that merely a fraction of a complex reality is found in the newspapers. This leads to a distortion of the real world, whereby the readers create a new reality based on the information they have at their disposal. Public opinion can be influenced by means of interpersonal communication about the topics presented by newspapers. (see Schulz, 2002: 355ff)

4 Content analysis

The most common way in science to systematically analyse data out of some text this to work with content analysis. Quantitative content analyses were first documented in the 18 century, so the tradition of using content analysis is quite long. Furthermore, content analyses enable researchers to analyse data without any interactive impact. Even large amounts of data from the past can be analysed this way. Researchers have the opportunity to analyse social and cultural values and their change. The main goal of a content analysis, therefore, is the systematic elaboration of any form of content. Although content analysis offers some advantages, this empirical method is used mostly when all others fail. (see Früh, 1998: 47; Krippendorff, 1980: 13f)

In addition to the choice of material which is to be analysed with some content analysis, the codebook is a crucial factor in the quality of the research. The codebook includes all variables which determine the collated data. Within this study these variables should collect formal attributes as well as causes of delinquency and measures presented in the media. As a result, some common criminological theories and approaches such as Merton’s theory of anomie, Hirschi’s social bonding theory or Cohen’s theory of subcultures were used in this work. (see Früh, 1998: 132ff)

Common criticism of content analysis speaks of lacking objectivity and reliability. Krippendorff mentions that “ […] the term reliability covers at least three distinct types and it would be better to designate which of these applies in a given situation.” (Krippendorff, 1980: 130) He differentiates between stability, reproducibility and accuracy. It is interesting to note that methodical errors occur due to individual interpretations. To reduce these errors, the study features pretests and reliability tests. (see Krippendorff, 1980: 130f) During this study the intercoder reliability was calculated twice with values of 0.96 the first time and 0.91 the second time.

Criminal and Antisocial Behaviour in Upper Austrian Print Media 134

5 Results

5.1 Overview

The prime objective of this content analysis was to describe the reporting of criminal and antisocial behaviour in Upper Austrian print media. Six dimensions were created to ensure that the analysis is comprehensive. These dimensions cover the following research questions:

• Design and format: What is the average format of an article and how are the articles laid out?

• Neutrality and objectivity: Are uncommon crimes presented in a more spectacular way?

• Crimes: Which crimes are presented by the newspapers? Can reality be found in the reporting?

• Offenders: Is there a focus on typical formal attributes among the offenders?

• Victims: What image of the victims do the newspapers convey? • Causes of criminal and antisocial behaviour and measures: According to

the newspapers, what are the causes of and the reasons for criminal and antisocial behaviour? Do the newspapers offer measures to prevent crime?

5.2 Design and format of the articles

A quick look at Table 1 shows that the majority of the articles cover a small area. Small articles were defined as the size of a credit card or smaller. All in all, this category includes more than 50% of all articles, whereas only 10% of the articles cover a large area. Primarily the free newspapers refrain from detailed reporting, whereas daily newspapers try to inform their readers with a more detailed approach.

An analysis of the headlines shows us a different picture. As headlines generally have the function of an eye-catcher, headlines in particular have more presence. An increase from small to large can be observed in every kind of newspaper. At nearly 25%, the total number of large headlines is higher than the corresponding number of large articles.
Table 2: Size of articles and headlines Article in % Headline in % Newspaper

small medium large small medium large Daily newspapers 50.3* 34.9 14.8 37.7 37.7* 24.6 Weekly newspapers 58.6 28.6 12.8 45.5 29.5 25.0 Free newspapers 74.2*** 23.1** 2.7*** 63.4*** 11.3*** 25.3 Total 53.6 33.2 13.2 41.2 34.1 24.7

n article = 1535; n headline = 1526; * p<0.05, ** p<0.01, *** p<0.001

Criminal and Antisocial Behaviour in Upper Austrian Print Media 135

4In about one-third of the articles at least one picture is used to highlight an article. Friedrichsen (2004) emphasises the power of pictures in creating emotions. A correlation between the number of pictures and the size of the article (r = .50***) and the headline (r = .34***) shows significant results.

5.3 Neutrality and objectivity

One of the most important principles of writing an article is to be as objective as possible. Golser (2006), however, pointed out that only 32% of German journalists decide to write without any bias, whereas the majority mentions the right to form their own interpretation. (see Golser, 2006: 43f)

The analysis of this study showed that all in all the statement of Golser (2006) is found in Austrian journalism too. About a third of the articles (35.1%) and headlines (31.5%) lacked the required objectivity. This missing objectivity is found in every sort of newspaper, whereby free newspapers in particular have significantly lower values. Furthermore, it is interesting to note that the length of an article has a significant influence on objectivity (r = .41***).

The following table gives us an overview of how strong the connection between certain crimes or antisocial behaviour and missing objectivity is. The table distinguishes between headlines and articles.