Sie sind auf Seite 1von 24

Developing and Validating a Bullying Bystander Emotion Reaction Scale (BBER) with Two Stages of Factor Analysis Process

Chiao-Lin Huang, Tsai-Wei Huang, I-Chi Tsao National Chiayi University 08/11/2012

Introduction
Bullying misbehaviors increased in school
80% of kids in Taiwan had witnessed various bullying in schools
The Child Welfare League Foundation in Taiwan (CWLFT, 2004)

Whats bulling?
An intentionally or deliberately negative action or behavior that has provoked or hurt others in the form of physical, verbal or other indirect methods (Olweus, 2003).

Bullying Behaviors (Olweus, 2003)


Physical Verbal Relational Counterattacking Sexual

Types of bulling characters


Bullies Victims Bystanders

Innocent for bystanders ?


Bystanders are not the innocents for they act in the character of helpers or facilitators through their silent acceptance or ignorance. (Coloroso, 2005; Liepe-Levinson, & Levinson, 2005 )

Bystanders Impact
Strengthen bias of the bullying event Differential treatment Hinder sympathy, mercy, and respect
(OConnell, Pepler & Craig, 1999; Coloroso, 2005; Salmivalli, 1999)

~the three basic elements of peer relationship

Bystanders emotion reactions were rarely discussed


(Boulton, 1995; Chiu, 2001, Coloroso, 2005; Hsu, 1999; Li, 2005; Rigby, 1996 ;Tseng, 2004; Wei, 2003) .

Bully-Victim relationships are spotlighted Rare researches discussed the practical impact of bullying on bystanders, especially on their emotion reactions. Lack of instrument for investigating bullying bystanders emotion reactions

Purposes
To development an instrument for investigating bullying bystanders emotion reactions (BBER)

Questions
1) Whats the contents of bullying bystander emotion reactions (BBER)? 2) How well is the quality of the BBER scale?

Method
Subject: 700 valid 5th and 6th grades (83.33% return rate) from 18 public elementary schools in Taiwan. 351 boys (50.14%) and 349 girls (49.86%).

Rationales of BBER
Wrath Sadness Fear Happiness Like Shame
(Goleman,2000; Heinrichs, 2003; Liepe-Levinson & Levinson, 2005)

Test construction
Rationales of BBER Experts opinions 6 dimensions (Wrath, Sadness, Fear, Happiness, Like, 27 items constructed.

Shame )

Item analysis by examining


Missing value Homogeneity Internal consistency Discrimination

Two stages of Factor Analysis


1st stage 2nd stage
Exploratory Factor Analysis (EFA) (N =176 ) Confirmatory Factor Analysis (CFA) (N =700).

Model fit evaluation

Computer software package


SPSS15 and AMOS5.0

RMSEA (Browne & Cudeck, 1993) CFI (Bentler, 1990) IFI (Bollen, 1989 ) PGFI (Mulaik, et al., 1989),.

RESULTS
Final 20 items remained by item analysis
Dimension Wrath 3 Sadness 2 Fear 5 Happiness 4 Like Shame 2 4 Subscale Mean 1.60 5.20 11.21 9.24 3.38 12.09 Subscale Standard Deviation 3.21 1.91 5.24 4.27 2.05 4.11 Single Item Mean 3.53 (1st) 2.6 (3rd) 2.24 (5 th) 2.31 (4 th) 1.69 (6 th) 3.02 (2nd)

Correlations between dimensions


Wrath Wrath Sadness Shame
Happiness

Sadness

Shame

Happiness

Like

Fear

1.00 .21 .39 (.22 (.44

, .19 , .34

) )

1.00 .12 .14 .18 (.10, .50 (.13 , .67

) ) ) )

1.00 .12 .15 .23 (.14 (.23 ( .18 , .09 ) , .03 ) , .30 )

.01 (.13 , -.09 ) -.02 (.06 , -.07) .33 (.39

1.00 .58 .36 (.60 ( .44 , .49 , .35

Like Fear
*p

( .13 , .20

) )

1.00 .28 ( .33 , .28 )


1.00

, .25

.35

( .37

, .20

.05; **p

.01. (351 boys, 349 girls)

Second-order factor model of BBER

2( 162 ) =370.73 (p < .001); RMSEA = .08; CFI = .80; PGFI = .63

Scale quality

Conclusion & Discussion


1.BBER comprise 6 dimensions (Wrath, Sadness, Fear, Happiness, Like, Shame ) with final 20 items remained. 2.Most bullying bystander students feel wrath, shame, and sad; few feel happy. 3.Low positive correlations existed among emotions of wrath, shame, and fear as well as those of happiness and like.

4.More than 75% values of factor loadings from EFA were greater than .50 with accompanying a Cronbach alpha coefficient of .83 and 64% variances explained by the 6 factors of BBER. 5.A second- o rder CFA model of BBER could be accepted with moderate model fit evidence. 6.Although only one of the average variance extracted by latent factors approached the .50 criterion suggested by Bagozzi & Yi (1988), the remains were still around .40. 7.Only the Sadness subscale was with composite reliability lower than .60 (Bagozzi &Yi,1988), those of other five subscales were greater than the criterion.

Limitations and Suggestions


Sample were not heterogeneous enough to optimize ideal alpha coefficients in subscales by EFA. Future studies might collect various data sets from different schooling levels of students and compare their differences on BBER scale. Future studies might apply the BBER to the bullying related topics.

Thanks for your listening!