Preliminarna Študija Vprašalnika Agresivnega Vedenja

MEASUREMENT OF STUDENTS' AGGRESSIVE BEHAVIOUR IN SCHOOL SETTINGS

Ana Kozina, Educational Research Institute, Slovenia [email protected]

Paper presented at the European Conference on Educational Research, University of Ghent, 19-21 September 2007

ABSTRACT

Aggressive behaviour refers to behaviour that is intended to cause harm or pain. It can be external or internal, direct or indirect. A negative influence on society in general is obvious, but the influence on school work should not be neglected as well. It seems that the problem of aggressive behaviour in school settings has been increasing through the years and therefore affecting school work at many different levels. Therefore it is essential to develop valid instruments for measuring aggression in the school settings in order to control aggressive behaviour and improve the school climate for effective learning. In the following study we present the development of psycho metrically valid instrument for measuring students’ aggression in school. We will use a newly developed instrument to identify correlates of student aggression according to different student and school factors. The students factors included in the analysis are: gender, age, educational aspirations, activities in spare time and attitudes toward school. The school factors are focused on the level of job satisfaction among teachers and their perception of school safety. The value of the new instrument is shown through its metric characteristics and through investigating the role and importance of aggressive behaviour in school. Its practical value is shown in identifications of the student and school factors that are associated with aggression.

Key words: aggression, children, school, instrument construction, metric characteristics

INTRODUCTION

The current study focuses on aggression, children and school. At the beginning it should be stated why these three. Well investigated and many times proven negative influence of aggression puts aggression in focus. Another point worth mentioning is that in trying to identify problems of modern school aggressive behaviour of students clearly represents one of them. While this already partly explains our focus on the school and children, there are still more explanations that need to be considered. One of them is that aggression is a stable personal trait lasting from childhood, through adolescence to adulthood (Loeber, Hay, 1997). As proven in many studies, it is a stable trait and when observed in children, it is a good predictor of later adult criminal behaviour (Ferris, 1996; Carr, 1998; Fossati, Maffei, Acquarini and Ceglie, 2003; van Lier 2005). Therefore, the only sensible thing is to try to influence it in childhood by identifying more aggressive individuals and trying to modify their aggression. That is especially important since aggression predicts future social, psychological, behavioural and educational problems (Schwartz, Nakamoto, Hopmeyer Gorman, McKay, 2006; Crick, 2006). Since school is an important factor in the process of

- 1 - socialization, the aim is to find correlates of aggression in school and try to influence aggression through them. The correlates already identified will be discusses later on. This paper will present the theoretical background for developing a new instrument for measuring aggressive behaviour of children in the elementary school followed by presentation of its development, use and its correlates within the school setting. The more specific objectives are:

…to develop new instrument for measuring aggression in school; …to investigate age and gender differences in aggression; …to investigate the role of aggression in school;

THEORETICAL BACKGROUND

AGGRESSION

As many other psychological phenomena aggression is an extremely complex concept and therefore, the definition itself is not very easy to obtain. If we have to sum up some of the existing definitions, we can say that majority of them are focused on its intent, expression and factors influencing it. They also include causing harm or intention to cause harm to another or self (Lamovec, 1988). Complexity is also evident from a variety of different aggression types. It could be classified either by its intention or its orientation. In intention oriented classifications, we can find instrumental aggression (orientated towards accomplishing specific goal), constructive aggression (oriented towards construction), destructive aggression (orientated towards destruction, harm…) and frustration aggression (orientated towards releasing energy or frustration). Commonly used classification by its orientation is shown in the figure below.

aggression

orientated IN orientated OUT

indirect direct indirect direct (guilt, (psychosomatic (physical, (hostility, autodestruction, illneses; depression, verbal, anger, suicide...) alcoholism...) negativism...) suspition...)

Factors influencing aggression vary and can be put into several different groups. -Biochemical factors: testosterone, male gender, low cholesterol, low nitro oxide, low serotonin, increased ER-beta, ER-alpha, dysfunctional amygdale, frontal lobe leisure… (Archer, 1991) -Chemical factors: ethanol, methamphetamine, alcohol, cocaine, marihuana, lead, apomorphin, fluoxetine and caffeine. -Psychological factors: low arousal, brain dysfunctions, anxiety, hyperactivity, early childhood aggression, early antisocial behaviour, positive attitudes towards aggressive behaviour an low IQ are found to be associated with higher aggression.

- 2 - -Sociological factors: transformation of sex roles in females, competition, other's aggressive behaviour, failure, parental criminality, neglect, abuse, indulged or strict discipline, inconsistent punishment, low parental involvement, poor family bonding, parents’ support of punishment, frequent residential moves, parent-child separation, educational failure, low bonding to school and dropping out of school. In general theories explaining aggressive behaviour put different emphasis on different aspects such as characteristics of individual (temperament, learnt behaviour, experiences of aggressive behaviour, physical illness…), characteristics of the environment (frustration, threats, stress, social dynamics…), institutions (school work, bad working conditions...) and cultural influences (social understanding of aggressive behaviour, diametrical expectations, structural unemployment, poverty) (Krall, 2003). The first in a row of theories that shaped the construct of aggression was instinctual theory, which claims aggression to be inherent in human behaviour. This theory was followed by frustration – aggression theory in 1939 developed by Dollard and Miller and explains that people are likely to become aggressive when they are frustrated. The emphasis on learning was introduces by Bandura and his Social learning theory developed in 1973 (Williams, Boyd, Cascardi and Poythress, 1996). Science today has accepted the complexity of aggression and is trying to integrate different backgrounds into one stable construct. Instead of partial theories explaining one or two aspects current science is interested in different integrative approaches. For example: aggression as a way of coping with stress, aggression as a masculine way of controlling environment, aggression as a result of productive coping with reality, and aggression as a consequence of economical poverty (Krall 2003). Aggression as a way of coping with stress evolves on the basis of unsuccessful conflict solving and is, therefore, a result of coexistence of individual and environmental factors. When a conflict appears, the first to follow is the evaluation of importance of the conflict for an individual involved. This evaluation is followed by a decision, regarding the conflict solving strategy. When certain conflict is evaluated as an important one and when strategies for solving the conflict are limited, the probability of aggressive behaviour increases (Krall, 2003). The other example is aggression as a result of socio – environmental conditions. This approach shows another way of interaction with environment by taking into account the social environment. It focuses mainly on the meaning individual assigns to himself and his environment.

AGGRESSION IN SCHOOL SOUROUNDINGS

Most common types of aggression observed in school settings are physical, verbal and psychological aggression (Popp, 2003). These types are commonly observed, especially physical aggression, due to the ease of spotting. In majority of the studies, the most frequent aggression found in school surroundings was physical and verbal aggression. By verbal aggression, we understand different forms of calling names, nicknames, and spreading rumours that lead to social stigmatism and exclusion. A study in Germany showed the presence of physical violence in schools (vandalism, fights, blackmailing) and also presence of verbal aggression especially in higher grades. The result of the study mentioned show that 50 to 60 % of children showed verbal aggression, while physical aggression was reported only by 2 % of children. But here we have to take under consideration that verbal and psychological aggression have longer lasting consequences and these forms of aggressions are more difficult to spot (Popp, 2003). There are of course all kinds of aggression found in the school surroundings but physical and verbal aggression

- 3 - are most common and nevertheless, easier to observe. A variety of different forms of aggression found in schools is shown in the table below.

Table 1. Classification of aggressive behaviour in school. latent aggression manifest aggression being angry, threatening, towards class torture, blackmail, attacking, attacking taking things... mates with weapons, robbing… yelling at, damaging towards teachers attacking, hurting, threats with knife, clothes, skipping classes... stocking, phone aggression, threatening letters... spraying, sketching, towards objects destroying furniture, windows, eliciting destroying flowers, fire, damaging cars… throwing bottles, slamming doors, damaging… verbal and physical sexual aggression rape harassments ... pulling hair out, scratching auto aggression suicide face… (Belser, 1999; Krall 2003)

A commonly shared belief strongly influenced by frequent media reporting is that aggressive behaviour in schools is increasing. This hypothesis, however, still needs to be proven. The study conducted in Germany, for example, shows that this is not the case and that between 1994 and 1999 aggression had not increased at all (Fuchs, 2001, Popp, 2003). Another question still waiting to be answered refers to different factors connected to aggression in school. So far, there has been evidence of important connections between aggression and grades, teacher – student relations, identification with learning material and the school climate in general (Krall, 2003). The influence of the grades is based on high parental expectations. Not reaching these high standards results in conflicts and aggressive behaviour (Popp, 2003, Tomori, 2000, Dekleva, 2000). As far as the teacher – student relations are concerned, the characteristics leading into more aggressive behaviour are rigid behaviour of a teacher and the use of institutionalized ways of power. When we consider the social climate in a classroom, the emphasis is put on connectedness among students, friendly relations and absence of unhealthy competitiveness (Popp, 2003). Important differences in aggressive behaviour were established among different types of schools (Krall, 2003).

MEASUREMENTS OF AGGRESSION

When addressing aggression in the school all conclusions must be result of valid, reliable and objective measures of aggression. The method of projection was commonly used in the past. It still is in the use today although the general trend is moving towards the use of self evaluation questionnaires and within school settings also towards teachers’ and peers’ evaluation questionnaire. Questionnaires and scales have a lot of advantages such as objectivity, validity, reliability, easy usage and so forth. Their most frequent criticism refers to giving false, usually socially more acceptable answers, which is a common case with aggression due the perception as aggression being negative trait. This could be avoided by including the so called scales for

- 4 - measuring the need for social acceptance. The most frequently used measure for aggression is Buss Durkee Hostility inventory (1957). It measures seven specific types of aggression: assault, indirect aggression, irritability, negativism, resentment, suspicion and verbal aggression. Researchers can therefore discover not only how aggressive person is, but also how this aggression is manifested. Our instrument belongs to the group of self evaluation measures. In attempt of getting as honest responses as possible it is going to be administered anonymously.

INSTRUMENT DEVELOPMENT

LA (Lestvica Agresivnosti) Aggression Questionnaire was developed in need of valid, reliable and objective instrument for measuring different types of aggression in school. The aim was to develop a new instrument, specifically constructed for and adapted to school use.

The item construction was based on detailed review of the literature and types of aggressive behaviour found in other studies on aggression in elementary schools. The goal was to include as many different types of aggression as possible (verbal, indirect, physical, suspiciousness, internal, auto aggression, anger, hostility, negativism…) and direct them towards different objects important in the school environment (class mates, teachers, selves, objects, parents…). Following this empirical plan, 96 items were set to compose first version of LA aggression questionnaire. All items were self evaluating, meaning that students reported about their own aggression. Afterwards, the first step was to determine metrical characteristics of this newly developed instrument in order to use it in school and to identify correlates of aggression. In order to achieve our goals, the study was organized in two parts; the first one being a preliminary study on a convenience sample and the second one being the main study on representative sample for Slovenia.

PRELIMINARY STUDY

A preliminary study of LA Aggression Questionnaire was conducted on a convenience sample of 162 (90 male and 79 female; average age in months was 143, 61 with SD = 23,062) students in elementary school and the first version of LA with 96 items was used. Responses were given on seven point scale (1= strongly disagree, 2=disagree, 3= mostly disagree, 4= something in- between, 5= mostly agree, 6= agree, 7= strongly agree). The items were positively evaluated with two exceptions (recoded in the later stage), meaning that the higher score stands for the higher level of aggression. Data for the preliminary study was gathered in 2 elementary schools in Slovenia in September 2005. The time for giving responses was not limited and majority of questionnaires were completed in 15 minutes. The purpose of the preliminary study was to extract items with good metric characteristics and to include them in second version of LA aggression questionnaire and therefore; descriptive statistics, structure analyses and reliability analyses were conducted. Due to a low standard deviation indicating low level of differentiation among students at different levels of aggression three items were eliminated. The rest of the item set was included in the component analysis for investigating the structure of the new instrument. A prior test for analyzing sufficiency of data for principal component analyses showed that data were suitable for analyzing structure (Kaiser – Meyer – Olkin = 0,668; Bartlett’s

- 5 - test of sphericity: χ2 (4560) = 7900, 191; p = , 000). With taking Kaiser Guttmann criteria (eigen value over one) under account 6 components explaining 42,431 % of the total variance were found. Items that were strongly correlated to these six components and that were logical in terms of the content were kept. This resulted in six components: anger/hostility, (16 items), verbal aggression (15 items), physical aggression (12 items), internal aggression (14 items), aggression towards authority (5 items) and suspiciousness (3 items). Cronbachs’Alpha as a measure of internal consistency was used as a method of reliability. The results showed sufficient reliability for the first five scales (α1=,845; α2=,864); α3= 825; α4=,799; α5=,786) and not for the sixth component which was eliminated. Reliability analysis conducted on item levels resulted in the single items being dropped from anger/hostility and aggression towards authority components. With the High and low groups method differentiation on item level was analysed. Students were divided into three different groups (with low, medium and high level of aggression) and differences on sole items between these groups were calculated. The outcome was an elimination of two items in the internal aggression scale. The result of all these analyses was the second version of LA aggression questionnaire with 58 items.

MAIN STUDY

In the main study the second version of LA Aggression Questionnaire with 58 self evaluation items was administered with a three different versions with rotated items in order to control the effect of tiredness. Scale was transformed into Likert scale response format (1= strongly disagree, 2= disagree, 3= something in between, 4= agree, 5= strongly agree). Three LA was administered together with TIMSS Field Test Questionnaire in March 2006 on a representative sample of 2780 fourth and eight grade students in Slovenia, participating in TIMSS 2007 Field Test Study. TIMSS (Trends in International Math and Science Study) is conducted by IEA (Evaluation of Education Achievement) and assesses achievements in the mathematics and science in the fourth and eight grades and collects a rich array of background information. TIMSS students’ questionnaires for the fourth and eight grade were used to investigate the role of aggression in school. Therefore, the TIMSS National Field Test Data Base was used to link of the school and students factors to the factors of LA. The data on students' gender, age, use of spare time, educational aspirations and attitudes towards school were used. The data on teachers’ perception of job satisfaction and school safety were also used from the TIMSS teachers’ questionnaire for the fourth grade and eight grade math and science teachers. The first goal was to analyse the structure of the instrument in order to find out weather it remains the same as in the preliminary study. A prior test for analyzing sufficiency of data for principal component analyze showed that data were suitable for analyzing structure (Kaiser – Meyer – Olkin = 0,971; Bartlett’s test of sphericity: χ2 (1891) = 51973,446; p =,000).

Table 2. Outcomes of exploratory factor analyses of second version LA Aggression Questionnaire. eigen value % of variance cumulative % 1 19,780 31,903 31,903 2 3,471 5,598 37,501 3 2,074 3,346 40,847 4 1,700 2,742 43,589

- 6 - 5 1,613 2,602 46,191 Note. As a method of extraction principal component analyses was used. The results were consistent with the preliminary study, again showing the structure of the six components (Kaiser Guttmann criteria) and again the sixth component was left out. On a more representative sample, a higher percent of the total variance is explained with the first five components. These five components formed a foundation for five scales or five types of aggression measured by the new instrument, named: anger/hostility, verbal aggression, physical aggression, internal aggression and aggression towards authority. Certain parts of this structure are consistent with other questionnaires measuring aggression (Buss, Perry, 1992). Loadings on the first factor are high on all items and similar results were found also elsewhere (Buss, Perry, 1992). This puts anger into focus. One explanation is that anger is psychological base for aggression. The fact is that people are more likely to act aggressively when they are angry. Correlates of anger and aggression are a result of other studies as well (Smith, Furlong, 1998; Campano, 2004).

20 m ean 15

0 anger/hostility verbal physical internal aggres sion aggres sion aggression aggression towards authority

Figure 1. Averages on different types of measured aggression in elementary school children.

Anger/hostility is the most frequent type of aggression in Slovenian elementary schools, followed by verbal aggression, internal aggression, physical aggression and aggression towards authority being the last one. The results are not consistent with the findings of other studies which show physical aggression as the most frequent one. The reason could be that the other studies in a way neglected more internal types of aggression because it is more difficult to notice them in school. Given that LA is a newly developed measure, prior any further use additional studies are needed to provide evidence of validity and reliability.

USE OF LA AGGRESSION SCALE

An additional aim in applying the new instrument to a representative sample of Slovenian school children was to test gender and age differences in aggression, and furthermore, to analyse relations

- 7 - between aggression and the use of spare time, students’ aspirations, school attitudes, teachers’ job satisfaction and teachers’ perceptions of school safety. The results for each of these objectives are discussed below.

GENDER

The majority of studies concerned with aggression show significant gender differences indicating that men are more aggressive. These differences were explained by using a variety of different approaches. Biological, social and evolutionary viewpoints were commonly used. In current modern science researchers from different fields are pointing out that gender differences exist due to different ways of expressing aggression. In this context, men usually use more direct forms of aggressive behaviour and women more indirect forms (Condon, Morales-Vives, Ferrando, Vigil- Colet, 2006). The same trends of gender differences were found in German studies (Holtappels 1997, Tilmann, 1999; Fuchs, 2001, after Popp, 2003) indicating that boys took precedence in physical aggression and girls in verbal aggression (Popp, 2003).

40 35 female 30 25 male 20 15 10 5 0 anger/hostility verbal aggression physical aggression internal aggression aggression towards authority

Figure 2. Gender differences in types of measured aggression.

Table 3. Outcomes of one way ANOVA for testing gender differences. scale F p anger/hostility 11,974 ,000 verbal aggression 20,304 ,000 physical aggression 92,357 ,000 internal aggression 3,742 ,024 aggression toward authority 24,210 ,000 Note. Statistically significant differences on ,05 level are bolded.

As expected gender differences are significant on all the LA scales. These types of findings are consistent with the studies presented above, yet, there is an interesting switch. Boys are found to be more aggressive on all scales, not only on physical aggression. Therefore, an explanation that only the expression is gender specific is not applicable to this data set. The results are consistent with the studies claiming that men are showing more aggressive behaviour and more externalizing problems (Buss, Perry, 1992; Delfos, 2004). The precedence of male gender in a level of aggression is not unexpected when taking other clinical problems and disorders into consideration (Diagnostic and Statistical Manual of Mental Disorders, American Psychiatric Association, 1994).

- 8 - We can see that boys take precedence in many other disorders as well (ADHD, autism spectrum disorder and others).

For testing the age differences the fourth and eight grade students were compared and additionally one way ANOVA was used to test the statistical significance of mean differences.

30 fourth grade 20 eight grade 10

0 anger/hostility verbal aggression physical aggression internal aggression aggression towards authority

Figure 3. Differences of level of measured aggression in fourth and eight grade.

Table 4. Outcomes of one way ANOVA for testing age differences. scale F p anger/hostility 156,665 ,000 verbal aggression 448,184 ,000 physical aggression 92,222 ,000 internal aggression 49,071 ,000 aggression towards authority 704,883 ,000 Note. Statistically significant differences on ,01 level are bolded.

Longitudinal studies of children are typically showing decrease in measured aggression from early childhood through adolescence (Cairns and others, 1989). The frequency of physical aggression in particular has been shown to decrease from early childhood on (Cairns, Cairns, Neckerman, Ferguson, Gariepy, 1989; Nagin, Tremblay, 1999; Tremblay 2000; Romano, 2005). The opposite trend is in verbal and internal aggression which appear to be increasing from childhood through adolescence (Romano, 2005). Our results are partly consistent with the literature. The increase in internal and verbal aggression is consistent with the literature, while the increase in physical aggression is exactly the opposite of what was expected. One of the hypotheses that still needs further research is that aggression itself does not increase, but because of the increased capacity for meta cognition, older children are more able to report their inner states and acts. In this data set, we have to point out that the study was not designed to be longitudinal thereby the differences could be an outcome of other fourth and eight grade population differences rather than age itself.

EDUCATIONAL ASPIRATION

Educational aspiration is a goal or intention of someone to reach a certain educational level. The level of students’ educational aspiration was tested in the eight grade with the question in TIMSS

- 9 - questionnaire: How far in school do you expect to go (secondary school; professional education after secondary school; 2-3 year university study (called higher study); 4 year university study; master or PhD)?

anger/hostility 8 verbal aggression physical aggression 6 internal aggression aggression towards authority 4

0 secondary vocational 2-3 year study 4 year study master or pHd school course Figure 4. Relation between educational aspiration and different types of measured aggression.

Table 5. Outcomes of one way ANOVA for testing differences in aggression and level of aspiration at eight grade students. scale F p anger/hostility 4,254 ,000 internal aggression 5,071 ,000 verbal aggression 9,065 ,000 physical aggression 2,586 ,017 aggression towards authority 4,935 ,000 Note. Statistically significant differences on ,05 level are bolded.

The differences among groups of students with different educational aspiration in measured aggression are significant on all the types of aggression. The higher the aspirations, the lower the level of physical aggression is found. The decrease in evident in all the types of aggression.

SPARE TIME

The spare time of students was assessed with the question in TIMSS questionnaire: On a normal school day, how much time you spend before or after school doing each of these things. Students responded to each of the possibilities (I watch television and videos; I play computer games; I play or talk with friends; I do jobs at home; I work at paid job; I play sport; I read a book for enjoyment; I use internet; I do homework) on a 5 point scale (no time; less then an hour; 1-2 hours; more the 2 but less the 4 hours; 4 or more hours). Differences were analysed with one way ANOVA separately for fourth and eight grade.

- 10 - Table 6. Outcomes of one way ANOVA for testing differences in aggression and use of spare time of fourth grade students. SCALE ACTIVITY anger/hostility verbal physical internal aggression aggression aggression aggression towards authority F (p) F (p) F (p) F (p) F (p) TV or videos 10,273 (,000) 9,171 (,000) 12,303 (,000) 4,827 (,001) 4,866 (,001) computer games 15,330 (,000) 12,696 (,000) 17,679 (,000) 1,842 (,118) 10,142 (,000) play with friends 1,209 (,305) 1,798 (,127) 2,037 (,087) 1,301 (,268) 3,121 (,014) do jobs at home 3,816 (,004) 5,127 (,000) 4,219 (,002) 2,141 (,074) 1,776 (,131) play sports 2,224 (,064) 2,346 (,053) 5,478 (,000) 0,548 (,701) 1,655 (,158) read book 1,588 (,175) 1,698 (,148) 1,716 (,144) 1,077 (,366) 4,145 (,002) use internet 5,944 (,000) 4,805 (,001) 6,555 (,000) 0,722 (,577) 6,449 (,000) do homework 0,812 (,517) 3,932 (,004) 2,636 (,033) ,652 (,652) 4,846 (,001) Note. Statistically significant differences on ,05 level are bolded.

Table 7. Outcomes of one way ANOVA for testing differences in aggression and use of spare time of eight grade students. SCALE ACTIVITY anger/hostility verbal physical internal aggression aggression aggression aggression towards authority F (p) F (p) F (p) F (p) F (p) TV or videos 6,861 (,000) 6,712 (,000) 9,089(,000) 5,247 (,000) 2,629 (,033) computer games 8,446 (,000) 6,311 (,000) 27,273 (,000) 2,667 (,031) 3,872 (,004) play with friends 0,708 (,587) 1,082 (,364) 1,052 (,379) 1,390 (,235) 3,702 (,005) do jobs at home 3,068 (,016) 9,522 (,000) 4,950 (,001) 0,949 (,435) 15,244 (,000) paid job 13,265 (,000) 1,961 (,098) 21,229 (,000) 3,892 (,004) 4,880 (,001) play sports 0,808 (,520) 0,604 (,660) 2,137 (,074) 1,514 (,196) 1,037 (,387) read book 8,885 (,000) 8,461 (,000) 13,109 (,000) 5,368 (,000) 6,402 (,000) use internet 7,256 (,000) 1,688 (,150) 6,068 (,000) 7,013 (,000) 12,543 (,000) Note. Statistically significant differences on ,05 level are bolded.

There are significant differences in the level of aggression among groups with different use of spare time. Differences are significant between groups that spent different amount of time watching TV or videos in both the fourth and eight grade, and playing computer games in the eight grade. More detailed look gives us relations that are completely consistent with findings of more than 1000 studies in the influence of watching TV on aggressive behaviour (Robinson, Wilde, Navracruz, Haydel, Varady, 2001). The relation indicates that individuals with higher levels of aggression spend more time watching TV and playing computer games. The relationship is not linear, with watching TV for no time at all being connected to higher anger/hostility then 1-2 hours or 2-4 hours. We have to point out that this anomaly is not present in both grades and not in all types of measured aggression and could be due to the small number of students that spent no time watching TV. As already mentioned above, there are a lot of studies concerned with the level of aggression and watching TV or playing computer games. Results of these studies also show that by

- 11 - reducing the time spent in front of TV and computer, aggression is reduced (Robinson, Wilde, Navracruz, Haydel, Varady, 2001). This appears to be one of the possible tools to reduce aggressive behaviour by reorganizing the spare time. Exactly the same trend is evident in the use of computers. The more time students spend on playing computers, the higher levels of aggression are present. Again, students who are not using computers at all are found to be more aggressive then those using computer for less then an hour a day. The relation is seen in both grades and in all measured types of aggression. The same, yet, more linear relations are shown with the use of the internet. The students who spent more time on the internet are found to be more aggressive. The results of Slovenian study (Gril, 2006) investigating use of spare time show precedence of passive activities, for example: watching TV, playing computer games… Therefore, the impact of these activities have on level of aggression is worrying and deserves special attention. The same trend of increase in aggression is seen in a paid job in the eight grade; the more time spent on the paid job, the higher levels of aggression are present. And here again relations are more complex; a student who spent 4 or more hours working on a paid job is evaluated as less aggressive than a student who spent 2-4 hours on a paid job. The opposite picture is seen with relations between working on household jobs and different types of aggression. The highest levels of aggression are present among students who do not spent any time on jobs at home and are decreasing with more and more time spent on household jobs at home. Reading books for enjoyment and doing homework show more complex and non linear relations. High levels of aggression are present with students who spent no time reading books for enjoyment and those who spent 4 or more hours reading books with decreases in-between. The same trend is evident in the relationship between the time spent on homework and different types of aggression in the eight grade but not in the fourth grade. Spending time with friends is connected with aggression towards the authority. Again, we are facing a trend of increase; thee more time students spent with friends, the more aggression towards the authority is shown with an exception of those students who spent no time with friends. The aggression towards the authority of the students who spent no time with their friends is even more evident among the fourth grade students. Here, the level of this type of aggression is high at extreme points of spending no time at all or four or more hours. With playing sport only physical aggression of the fourth grade students is associated. The more time students spent on playing sports, the more physical aggression is present. The relations between aggression and activities in the spare time are evidently important, but extremely complex. Therefore, additional research is requested.

ATTITUDES TOWARDS SCHOOL

Attitudes towards school were evaluated by using three items of TIMSS questionnaire: I like being in school; I think that in my school students try their best and In our school teachers want students to do their best. Students marked their level of agreement with these statements on a 4 point scale (strongly agree, agree, disagree, and strongly disagree).

- 12 - Table 8. Outcomes of one way ANOVA for testing differences in aggression and students’ attitudes towards school. fourth grade anger/hostility verbal aggression physical internal aggression towards aggression aggression authority F p F p F p F p F p 1 3,030 ,000 3,171 ,000 1,163 ,000 2,357 ,000 6,618 ,000 2 1,328 ,057 2,099 ,000 1,577 ,010 1,777 ,001 2,487 ,001 3 1,621 ,003 2,056 ,000 1,230 ,147 1,737 ,002 2,580 ,000 eight grade 1 14,723 ,000 10,720 ,000 15,847 ,000 7,624 ,000 24,269 ,000 2 12,518 ,000 14,306 ,000 4,428 ,004 7,057 ,000 22,487 ,000 3 13,177 ,000 4,784 ,003 16,933 ,000 3,651 ,012 7,075 ,000 Notes. 1 – I like being in school; 2 - I think that in my school students try their best 3 - In our school teachers want students to do their best. Statistically significant differences on ,05 level are bolded.

The differences in measured aggression of groups with different attitudes towards school are significant almost on all three items, all types of aggression and in both grades. Results of the fourth grade students for the first item are presented below. The same trend is seen also in the eight grade and in the other two items.

30 anger/hostility

25 verbal aggression physical aggression 20 internal aggression

15 aggression towards authority

0 strongly agree agree disagree strongly disagree

Figure 5. Attitude towards school and different types of aggression (fourth grade).

The more positive the attitudes towards school, the lower the level of aggression is present in both grades and in all types of measured aggression. We can conclude that all the types of measured aggression and attitudes towards school are strongly connected.

- 13 - SCHOOL FACTORS

In order to investigate the influence of aggression on school work we tested the differences in aggression according to different school factors. For representatives of the school factors the teachers’ job satisfaction and their perception of school safety were chosen. These two were chosen on assumption that they are significantly connected to aggression. We assumed that aggression of students would have an important influence on the level of job satisfaction of teachers, meaning that higher levels of aggressive behaviour would be connected with lower job satisfaction and also with lower perception of school safety.

Teachers of students involved in the TIMSS study evaluated teachers job satisfaction on their school on 5 point scale (very low, low, medium, high, very high). The results were compared to the levels of measured aggression.

Table 9. Outcomes of one way ANOVA for testing differences in aggression and teachers’ job satisfaction of eight and fourth grade students. fourth grade eight grade math teachers science teachers F p F p F p anger/hostility ,705 ,549 2,900 ,021 0,262 ,853 internal aggression ,112 ,953 1,951 ,100 3,627 ,013 verbal aggression ,273 ,845 3,056 ,016 1,556 ,198 physical aggression ,762 ,515 0,560 ,692 1,367 ,251 aggression towards authority 1,470 ,221 1,746 ,138 0,550 ,648 Note. Statistically significant differences on ,05 level are bolded.

The table shows few isolated significant differences in eight grade in the level of aggression concerning the teachers’ job satisfaction. We may conclude that students’ aggression does not have a direct impact on the teachers’ job satisfaction. The job satisfaction is a very complex phenomenon and, therefore, the influences are not straightforward. The differences found on our eight grade sample indicate that with the higher level of job satisfaction the lower level of anger/hostility, verbal and internal aggression is present. The reason behind it could be that since physical aggression and aggression towards authority are easily spotted by teachers and afterwards strongly sanctioned, and other forms are more obvious to the teachers and could, therefore, influence their job satisfaction most.

In addition, we conducted analyses of the school factors concerning safety at school. Again, we assumed that there are important connections between teachers’ evaluation of school safety and the level of students’ aggression. Teachers evaluated safety on following items This school is located in a safe neighbourhood; I feel safe at this school; This school’s security policies and practices are sufficient on 4 point scale (strongly agree; agree, disagree, strongly disagree).

- 14 - Table 10. Outcomes of one way ANOVA for testing differences in aggression and school safety. fourth grade safe neighbourhood feel safe security policies F p F p F p anger/hostility 9,569 ,000 5,204 ,006 2,621 ,073 verbal aggression 3,553 ,029 5,014 ,007 4,575 ,010 physical aggression 2,847 ,058 1,957 ,142 1,538 ,215 internal aggression 4,451 ,012 3,450 ,032 0,754 ,471 aggression towards 2,484 ,084 3,924 ,020 3,837 ,022 authority eight grade – math teachers anger/hostility 3,035 ,028 1,305 ,271 3,585 ,013 verbal aggression 2,237 ,082 1,831 ,140 3,466 ,016 physical aggression 1,443 ,229 ,995 ,395 2,459 ,061 internal aggression 2,895 ,034 2.254 ,080 0,320 ,811 aggression towards 1,901 ,128 4,563 ,003 2,534 ,055 authority eight grade – science teachers anger/hostility 0,581 ,559 0,805 ,447 0,239 ,869 verbal aggression 3,787 ,023 3,598 ,028 1,877 ,131 physical aggression 2,346 ,096 0,349 ,705 0,713 ,544 internal aggression 0,949 ,387 0,135 0,874 1,210 ,305 aggression towards 2,471 ,085 3,842 ,022 0,995 ,394 authority Note. Statistically significant differences on, 05 level are bolded.

We can see some important connections, especially with verbal aggression among the fourth (all three items) and also in the eight grade students (one and two items). A more detailed look gives us an interesting relationship explaining these connections. In the school that is evaluated as a safer one, students’ verbal aggression is lower than in other schools. The same pattern is seen in other important connections. Taking both evaluated school factors into account, we can conclude that in our data set verbal aggression is found to be more strongly connected to the school factors then other types of aggression measured by our instrument.

CONCLUSION

The study present the first step in constructing the new instrument with the purpose to identify aggressive students that need help in elementary school and prevent their future aggressive behaviour. We developed a psychometrically valid instrument designed specifically for measuring aggression in the school settings. In this area of research, AQ (Aggression questionnaire developed by Buss and Perry, 1992) is most frequently used. However, the reason that moved us towards developing our own instrument was that researchers using this questionnaire pointed out sensitivity to cultural and language differences. LA Aggression Questionnaire that we developed was found to be reliable and to differentiate efficiently among students of different levels of aggression. The paper presents first stages of the instrument development and, prior to any further use, the

- 15 - instrument should be tested on other samples for validity and reliability. The use of the LA aggression questionnaire in order to investigate correlates of aggression in schools should be understood as guidelines for possible further research. The results show that there are important connections of aggression and the use of spare time, attitudes toward school, and educational aspiration. Therefore further research should be done in this direction with analysing these relations in more detail. Not only that the study showed important connections on the student level but also in the teachers level as well. Taking into consideration that these are just few of many factors influencing aggression, there are still many more to identify. The main importance of the study is in presenting that by using the aggression questionnaire together with investigating other factors gives us a possible means to influence the level of aggression in children and to develop tools for decreasing aggression.

REFERENCES

Archer, J. (1991). The influence of testosterone on human aggression. British Journal of Psychology, 82 (1), 28.

Buss, A.H., Perry, M. (1992). The aggression questionnaire. Journal of personality and social psychology, 63 (3), 452-459.

Carr, A. (1998). The handbook of child and adolescent clinical psychology. London: Routledge.

Campano, J.P. (2004). Anger and aggression among Filipino students. Adolescence. 12 (22).

Crick, N.R. (2006). A longitudinal study of relational aggression, physical aggression, and children’s social-psychological adjustment. Journal of Abnormal Child Psychology, 4 (1).

Cairns, R.B., Cairns, B.D., Neckerman, H.J., Ferguson, L.L., Gariepy, J.L. (1989). Growth and aggression: 1. Childhood to early adolescence. Developmental psychology, 25 (2).320-330.

Condon, L., Morales-Vives, F., Ferrando, P.J., Vigil-Colet, A. (2006). Sex differences in the full and reduced versions of the aggression questionnaire. A question of differential item functioning? European Journal of Psychological Assessment, 22(2), 92-97.

Dekleva, B. (2000). Šola, mladi, nasilje. [School, youth and violence] V: Šelih, A. (ur), Prestopniško in odklonsko vedenje mladih – vzroki, pojavi, odzivanje. [Delinquent behaviour in youth origins, types and responses] Ljubljana: Bonex, str 137 – 149.

Delfos, M.F. (2004). Children and behavioural problems – anxiety, aggresion, depression and ADHD – a biopsychological model with guidelines for diagnostics and treatment. London: jessica Kingsley Publishers.

Ferris, C.F. (1996). Understanding aggressive behaviour in children. New York: The New York Academy of Science.

- 16 - Gril, A. (2006). Prosti čas mladih v Ljubljani. Ljubljana: Ped

Krall, H. (2003). Mladina in nasilje: teoretične koncepcije in perspektive pedagoškega ravnanja. Sodobna pedagogika, 54(120),2, 10-25.

Williams, T.Y., Boyd, J.C., Cascardi, M.A., Poythress, N. (1996). Factor Structure and Convergent Validity of the Aggression Questionnaire in an Offender Population. Psychological Assessment, 8 (4), 398 – 403.

Fossati, A., Maffei, C., Acquarini, E., DiCeglie, A. (2003). Multi group confirmatory component and factor analyses of the Italian version of the aggression questionnaire. European journal of psychological assessment, 19 (1), 54-65.

Lamovec T. (1988). Priročnik za psihologijo motivacije in emocij [Manual for Motivation and Emotion]. Ljubljana: FF, oddelek za psihologijo.

Loeber, R. in Hay, D. (1997). Key issues in the development of aggressoin and violence from childhood to early adulthood. Annual Review of Psychology,48, 371- 410.

Popp, U. (2003). Nasilje v šoli in koncepti njegovega preprečevanja. [Violence in School and its prevention] Sodobna pedagogika, 54(120),2, 26-41.

Robinson, T.N., Wilde, M.L., Navracruz, L.C. Haydel, K.F., Varady, A. (2001). Effects of Reducing Children's Television and Video Game Use on Aggressive Behaviour: A Randomized Controlled Trial. Archives of Pediatrics & Adolescent Medicine. 155(1):17-23.

Romano, E. (2005). Multilevel Correlates of Childhood Physical Aggression and pro social Behaviour. Journal of Abnormal Child Psychology, 10 (1).

Smith, D.C., Furlong, M.J. (1998). Addressing youth anger and aggression in school settings. Psychology in the schools, 35, 201-203.

Schwarzt, D., Nakamoto, J., Hopmeyer Gorman A., Mckay, T. (2006). Popularity, social acceptance, and aggression in adolescent peer groups:links with academic Peformance and school attendance. Developmental psychology,42(6), 1116-1127.

Tomori, M. (2000). Psihosocialni dejavniki pri mladostniškem prestopništvu. [Psychosocial factors of adolescence delinquency] V Šelih, A. (ur.), Prestopniško in odklonsko vedenje mladih – vzroki, pojavi, odzivanje. [Delinquent behaviour in youth origins, types and responses] Ljubljana: Bonex, str 89 – 111.

Van Lier, P.A.C. (2005). Trajectories of peer nominated aggression: risk status, predictons and outcomes. Journal of abnormal child psychology. 2(1).

- 17 -