Measuring gender attitudes in highly
patriarchal settings: developing measures in
Pakistan and Afghanistan
PROF RACHEL JEWKES
DIRECTOR, GENDER & HEALTH RESEARCH UNIT, SOUTH AFRICAN MEDICAL RESEARCH COUNCIL
CONSORTIUM DIRECTOR, WHAT WORKS TO END VIOLENCE
SECRETARY, SEXUAL VIOLENCE RESEARCH INITIATIVE
Standard measures of gender equity:
e.g. full scale is 34 items, GEM scale (short
version below as used in South Africa)
SECTION 2 ATTITUDES ABOUT RELATIONS BETWEEN MEN AND WOMEN
For each of the following statements please say answer whether you strongly agree, agree, disagree or strongly disagree with the following statements:
201
A
B
C
D
E
F
G
A woman’s most important role is to take care of her
home and cook for her family.
Men need sex more than women do.
There are times when a woman deserves to be beaten.
It is a woman’s responsibility to avoid getting pregnant.
STRONGLY
DISAGREE
DISAGREE
AGREE
STRONGLY
AGREE
1
2
3
4
1
2
3
4
1
2
3
1
2
3
A woman should tolerate violence in order to keep her
family together.
1
2
3
If someone insults a man, he should defend his
reputation, with force if he has to.
1
2
3
1
2
3
1
2
3
To be a man, a person needs to be tough
Some culturally
inappropriate
questions which
cannot be used
e.g 16 items on sex
and condoms

Once these are
removed the
scale isn’t great in
Asia, the internal
consistency is low,
especially for men
4
4
4
4
4
H
A man should have the final say in decisions in his home

4
Need to think differently about gender attitudes
in very patriarchal, conservative settings

Issues:

Constrained movement of women
including barriers to social participation

Limited space for women in household
decision-making

Acceptance of violence against
women

Some contestation over men’s role as
carers for women in the family

Process of translation of these issues
into questions which are appropriate
for use with children and adults

Core questions developed and used
in Afghanistan and Pakistan, with
some differences in extra questions

Constraints :

space in questionnaire/ time in
interviews
Core gender attitudes questions:
Pakistan
GENDER ATTITUDES (GA)
(Developed by Afghanistan team)
Circle the number that describes how much you agree or disagree with each
statement.
GA1
I think girls in our family should go to school
GA2
GA3
I think the husbands in your family should give permission to give their wives to go to
the clinic
I think the husbands in the family should listen to their wives’ opinion on schooling
GA4
I think the wives in the family should have a say in how money in their family is spent
GA5
I think the wives in the family should be able to ask a religious scholar about issues
GA6
I think the husbands in the family should respect the opinion of their wives on matters
related to income generating work
GA7
GA8
GA9
I think a husband in the family should be kind and caring toward the women in his
family
I think that the wives in our family should always obey their husbands
I think that if a wife in our family does something wrong her husband has the right to
punish her
Strongly
Agree
Agree
Disagree
Strongly
Disagree
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Women’s participation: questions also
used in Pakistan (822 boys & 930 girls, 6th grade)
Cronbach's alpha
(internal
consistency)
Women’s Participation (WP) (Developed by Afghanistan
team)
Circle the number that describes how much you
agree or disagree with each statement.
WP1
WP2
WP3
WP4
Strongly
Agree
Agree
Disagree
Strongly
Disagree
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
I think women should be
able to participate in
Weddings
Neighborhood events
Skills training (e.g. computer
skills, embroidery)
Income generating activities
PAKISTAN
Personal views (girls)
Personal views (boys)
0.78
0.73
Boy
Girl
10
20
30
genatt1
40
50
Gender attitudes comparing Afghanistan and Pakistan
90
80
70
60
50
40

Including 100 Afghan
women interview for pilot

Afghan women are more
conservative than
children in Pakistan

In Afghanistan: Lower
levels of agreement that
wives should obey their
husbands

Also on right of husbands
to punish wives
30
20
10
0
girls should husbands
husband
wives
go to school should give
should have
should
permission
a say in
listen to
to their
their wives’ how money
wives to go opinion on
is spent
to the clinic schooling
wives
should be
able to ask
a religious
scholar
about
religious
issues
WOMEN FOR WOMEN Strongly agree %
RIGHT TO PLAY: girls Strongly agree %
husbands
should
respect the
opinion of
their wives
on income
generating
work
husbands
wives
if a wife
should be
should
does
kind and always obey something
care about
their
wrong her
the
husbands
husband
happiness
has the
of women
right to
punish her
RIGHT TO PLAY: boys Strongly agree %
68
31
47
31
wives should always obey
their husband
54
a husband should be kind
and care about the
happiness of women in his…
39
41
husbands should respect
their wives on income
generating work
WOMEN FOR WOMEN Community view %
wives may ask a religious
scholar about religious issues
32.3
wives should have a say in
how money is spent
56
husbands should listen to
their wife on schooling
33
husbands should permit wife
to go to the clinic
80
70
60
50
40
30
20
10
0
girls should go to school
SOCIAL NORMS AND PERSONAL VIEWS:
AFGHANISTAN
WOMEN FOR WOMEN Personal view %
62.6
48.5
39.4
29
11.1
13.3

Considerable
difference between
perceived community
views (social norms)
and personal views

Social norms generally
more conservative
than personal views
Attitudes among women in two Afghan communities
40
35
30
25

All interviews done in Kabul
province but in Tajik and
Pashtun communities

Shown here are expanded
VAWG questions

Very substantial variation in
perceived social norms
around VAWG by ethnicity
20
15
10
5
0
In this
I think, if a wife
In this
I think, a wife
In this
I think, it is a
community, if a does something community, a who does things community, it is good thing for a
wife does
wrong her
woman who
that are wrong good to beat a young wife to be
something wrong husband has the does things should be beaten young women to beaten to teach
her husband has right to punish wrong should be to correct her teach her how to
her how to
the right to
her
beaten to correct
behaviour
behave properly behave properly
punish her
her behaviour
Pashtun community % agree
Tajik community %agree
Afghanistan: gender attitudes, ethnicity and
VAWG - reported attitudes quite different from violence
prevalence
Personal and community gender attitudes
by ethnic group (high=more gender
equitable)
Emotional and physical IPV in past 12
months by ethnic group
Pashtun
Pashtun
Tajik
Tajik
25
30
35
community atts
40
45
ithink
10
20
30
pasco
40
50
Multivariable logistic regression models of factors
associated with peer violence victimisation and
perpetration in/by girls and boys in 6th grade in Pakistan
GIRLS
PERPETRATION
BOYS
Low or no
High violence violence
perpetration perpetration p value
Patriarchal gender attitudes scale (scores)
Multi-variable logistic regression (other
variables not shown)
Patriarchal gender attitudes scale
Low or no
High violence violence
perpetration perpetration p value
10.2
9.05
0.001
10.6
odds
ratio
95% Confidence
Interval
P value
odds
ratio
1.10
1.05
1.15
GIRLS
9.36
95% Confidence
Interval
<0.0001 1.08 1.02
1.14
VICTIMISATION
BOYS
Low or no
High
violence violence
perpetration perpetration p value
0.005
9.88
P value
odds
ratio
0.004
1.15 1.07
8.64
95% Confidence
Interval
1.22
Low or no
High violence violence
perpetration perpetration
<0.0001
10.32
P value
odds
ratio
<0.0001
1.13
8.8
95% Confidence
Interval
1.03
1.24
p value
0.022
P value
0.01
Conclusions

New scales are fairly effective in Afghanistan and Pakistan and have quite
good statistical properties

Variations are as might be expected overall

The relationship between individually held gender attitudes and social
norms in Afghanistan is complex as ethnicity is much more important as a
predictor of experience of IPV than attitudes or norms

In Pakistan, personal gender attitudes are important as a factor associated
with peer violence perpetration and victimisation for girls and boys

This analysis is work in progress, especially as the main Afghan data is yet to
be collected
Acknowledgements

DFID for funding the research through What Works Global Programme

HTAC team: Mohamed Osman Hemat and Abdibullah Rahmatulluh

AKU team: Rozina Karmaliani, Hussain Khuwaja, Tazeen Ali, Yasmeen
Somani, Rozina Somani, Saleema Gulzar, Nargus Asad, Shireen Shezhad

Eureka/ WFWI team: Husain Ali, Mohamed Shafiq, Frozan Khan, Fazal
Karim, Carron Mann, Aishwarya Ratan

Julienne Corboz, What Works TA in Kabul

Andy Gibbs