Global Forum On Gender Statistics

Rome 10 -12 December 2007

Measuring violence against women: indicators and measurement tools

The Italian Women Safety Survey: methodological challenges and new achievements

Roberta Barletta, Isabella Corazziari, Alessandra Federici

Maria Giuseppina Muratore, Giovanna Tagliacozzo

Safety Women Survey

The Italian Women Safety Survey

The first Italian survey on violence against women

From a partnership with the Department for Rights and Equal Opportunities - Italian Presidency of the Council of Ministers

The first official numbers on violence against women (February 2007)Indicators on1. Prevalence, by different violence forms and different violent perpetrators2. Intensity3. Seriousness4. Consequences5. Seeking for help 6. …………..

Thanks to 25.000 women aged 16-70 years old, interviewed from January to October 2006

Thanks to shelters for women victims of violence

Thanks to availability of abused women

1. Which kind of information

There is the need for an integrated approach to violence gender based, to have accurate data that

meet users and community needs highlight the hidden violence phenomenon and contribute in

eliminating stereotypes define priorities for policy makers allow social and cultural change in combating violence against

women

But it’s necessary a

Balance between the opportunity to focus in depth on violence issue and the respondent burden

depending on the type of survey, if dedicated or a module

Focus on physical and sexual violence (partner and non partner) Focus on domestic violence

1. Psychological and verbal violence2. Economical violence3. Stalking

Focus on: Victim’s violence perception as a crime Seriousness (injuries and type of injuries, perceived seriousness,

feeling in danger of life, use of medicine and therapy to cope with violence)

Intensity (one or more times violence occurred, repetitiveness) Violence dynamics (arms, alcohol abuse of perpetrators) Reporting behaviour and women capacity of seeking for help

(shelters, with whom women speak of violence, police relationship...)

Children witness of violence Violence in pregnancy

2. Which kind of information: the essential issues

Risk factors of violence1. Abuse in WOMEN background

Mother abuse history Childhood victimisation

2. Abuse in the PARTNER background Experience of violence in childhood Witness of father violence against own mother’s

3. Individual factor risks partner’s related partner’s alcohol abuse Partner violent outside family too

4. Social factor risks partner related Women considered as an object to

denigration and berate

3. Which kind of information: the essential issues

1. How to collect data

Need of methodological and procedural dedicated tools to guarantee women safety, help women disclosure,

have sounded data

1. Appropriate setting

Risk to underestimate data if collected within a no specialized survey

(the life course rate of rape or attempted rape was 2,9% in Italian safety citizens’ survey (a victimization survey) against the 5% of violence survey (since the age of 16teen)

2. Accurate planning phase involving community: shelters, users, policy makers …

Pre-test Focus groups

Workers in shelters for women victims of violence Women victims of domestic violence Women from the community Interviewers who have already had experience in

victimisation surveys

Interviews to key professionals Police, legal and social experts

Pilot survey

2. How to collect data

Regarding the Content of the questionnaire and Procedural methods

3. No name violence Nor in the name survey Nor in the advanced letter Nor from the people answering to the toll free number Nor in the introduction Nor in the questions

Data are collected investigating behaviours The type of violence is defined in a way that women

can remember and reflect their lives as if in a mirror

The psychological violence questions are measured considering the different aspects of the daily life

3. How to collect data

For example: physical violence

ranked from the less to the most serious one:

threat to be physically hit to be pushed, grabbed to be yanked or knocked with an object to be slapped, kicked, punched or

bitten attempted strangulation, of a choking,

burning threats with weapons

4. Don’t be afraid to ask to women regarding their violent experience (motivate them with the importance of the study)

4. How to collect data

Different screening on partners and former partner are very important, the partner violence rates increase

Don’t hesitate in asking sensitive form of violence, sometimes the interview represents the only occasion to speak of violence

33% of women spoke of suffered partner violence for the first time with the interviewers, 45,2% in case of current partner violence

Women availability during the interview was mostly very good 52,7% and enough good 31,6%

For example: sexual violence

rape other form of rape (anal or oral penetration)

(only if no at rape question) attempted rape sexual intercourses with a third party undesired sexual intercourses, suffered

for fear of consequences degrading and humiliating sexual activities

(only for partner violence) other sexual violence forms not included before

5. Attention to graduate items and to the sequence of sections

5. How to collect data

Ask gradually about violence

The questions measuring violence should be included gradually

Ask questions on the every day life, leisure time, social networks, health, before those on violence

Insert the questions on violence from the partner in the section on the partner’s characteristics, after having asked about the relationship and the psychological violence battery

Test: the sequence of the sections

After the pilot survey

we tried to ask before the partner screening that the non partner one, but the results were no interesting:

no significant improvement in partner violence disclosure

a lower rate of non partner violence arose

So no changes were done in the full fledged survey

6. Attention to women safety

To choose the right methodologyThe telephone technique as a guarantee for anonymity, higher privacy, possibility to interrupt the telephone call in every moment, possibility to take easily an other appointment

Find the best time for her It’s important to interview the woman when she is alone at home

(with no partner in the house) Large timetable 9 a.m. -9 p.m. Possibility to be reached to a mobile phone To call back and to take an appointment

Reassure about privacy issues/anonymity

Create a good climate of confidence and faith

Toll free number

6. How to collect data

7. Attention to emotional trauma

Address to shelters

Female interviewers well recruited and well trained at the aim to be supportive but not a counsellor

Multidisciplinary approach of research team Psychologist Sociologist Statisticians Economist of organization

7. How to collect data

The Interviewers characteristics

Female; Minimum 24 years old; Comfortable discussing issues related to violence against

women; Sensitivity and maturity; Professional experience in CATI surveys as well as in

dealing with cases of violence (according to the type of the group);

Prior experience in handling similar sensitive research studies;

Listening skills, empathy, no counselling; Probing, no judgment; Warm tone of voice that helps creating a positive climate; Capacity to keep adequate detachment; Skills to elaborate own emotions and to handle unexpected

situations; Motivation

Interviewers Recruitment sheet

TOTAL SCORE

PREFERRING REQUIREMENTS NO YES

1.      Experience in social surveys by telephone

2.    Previous work and/or personal experiences underliningrelational, aid, service skills (teaching, help desk , voluntarywork, selling agent, theatre and expressive activities, phonecounselling, etc.)

3.     Previous working experience in administering interviews bypaper questionnaire

4.       Previous working or training experiences related tostatistical surveys

5. Documented experience lasted at least one year ofaid/assistance/counselling activities with women victims ofviolence in women shelters and/or telephone services foraid and assistance to women victims of violence

NOTES:

SPECIFY

INTERVIEWER First Name__________________________ Second Name_______________________

Interviewers Recruitment sheet DATA OF THE

INTERVIEW

Scarce/Low

Mediocre/Medium

Low

Enough/Medium

Good/Medium

High

Optimal/High

Extra-working activities related to the sectorMotivation to work as an interviewer

Dialect accent

Voice tone

Thinking ability on violence and survey themsConsciousness and control of emotions with respect to Consciousness and control of emotions of the victimPresence of stereotypes on victimised women

Training

NOTES

Selection consultant

INTERVIEWER

First Name ____________________________ Second Name________________CODE__________ Birth date_________________

Years of experience Whose of which as Title of study___________________________ working_____________________________

The pilot survey:Interviewers from shelters versus CATI professional interviewers

There are many differences among the two group of interviewers

1.Against every expectation, interviewers from shelters obtained more interviews than the other ones

2.They were “new” for this work, they don’t use dangerous automatism to maximize their work, each call is an adventure

3.They are able to conquer the household, to put at own ease the respondent, to value her

Refusal rate

Refusal of selected person

Interruption rate

Length

Interviewers from shelters

7,9 1,4 0,2 32

CATI professionals interviewers

10,1 1,7 1,5 26

The pilot survey:Interviewers from shelters versus CATI professional interviewers

0.5

2.1

3.2

0.5

0.3

4.8

0.4

1.8

4.5

13.5

0.4

9.0

10.9

11.2

1.3

5.8

0 2 4 6 8 10 12 14 16

Sexual violence by partner (last year)

Physical violence by partner (last year)

Sexual violence by partner

Physical violence by partner

Sexual violence by non partner (last year)(*)

Physical violence by non partner (last year)

Sexual violence by non partner(*)

Physical violence by non partner

Not from shelters From shelters

The interviewers from shelter: Read slowly and entirely the questions Gave time to reflect and to answer to

the interviewees create a confidant climate with women

This means more ability to capture violence

- Women disclosure themselves more –

They can be found not only in interviewers from shelter, but also in interviewers that directly experienced

violence in their life

Availibility

Sensitivity

Professionalism

Motivation

Are very important

The pilot survey:Interviewers from shelters versus CATI professional interviewers

What about the interviewers’ training

The continuous training: theoretical briefing practical exercitations technical briefing apprenticeship supported trial period

Methodology lectures exercises brainstorming and group discussion Audiovisual and video use role-playings

A questionnaire for emotiveness: emotions and wishes

Sharing/groups/feedback Anger/frustration/dejection Letting up of pression/organization of the working

environment Helping/intervening/restituting to interviewers Few emotions /boredom/detachment

What about the interviewers’ training: debriefing

8. Attention to representative data and accurate estimates

• Big sample size 25.000 interviews

8. How to collect data

Design: two stages random sample stratified at the first stage

First stage - households present on the official list of telephone

subscribers Stratification criterion

- Stratus variable: region and type of municipality Second stage

- Women aged 16-70 years old Selection criterion

- Random selection between eligible women

The main results: the figures of violence

6.743.000 women aged 16-70 have suffered physical or sexual abuse

31,9% of women 16-7018,8% physical abuse24,7% sexual abuse4,7% rape or attempted rape

14,3% by partner24,7% by non partner

Conclusion

Many things can be done to achieve good data,

taking into account the social and cultural context of each specific country

But some problems still remain above all regarding the use and the actual acceptance of the outcomes from the data, especially those regarding domestic violence

It’s then important also to work on the cultural context receiving data:

mass media workers, policy makers and general public