Children and women with disabilities

ANNEX I

RESEARCH METHODOLOGY FOR CONDUCTING KNOWLEDGE, ATTITUDES, BEHABIOUR, PRACTICE AND SOCIAL NORMS (KABPSN) STUDY ON

CHILDREN AND WOMEN WITH DISABILITIES

(Option 1)
Background

Since the fall of the Soviet Union, a number of local and international organizations have been working in Tajikistan in order to strengthen an inclusive society for persons with disabilities. The activities include capacity building trainings, awareness raising events and campaigns, lobbying and advocacy for the rights of persons with disabilities. Some positive changes have been made in service provision and integration of children with disabilities into regular schools, and making them more visible at community level through created playgrounds and resource centres. However, full acceptance of persons with disabilities by persons without disabilities will not occur until social barriers against the inclusion of persons with disabilities are eliminated. Such barriers include discriminatory attitudes, prejudices or stigma against children with disabilities (CwD) and women with disabilities (WwD) in the communities, and widely held negative beliefs towards disability, such as “disability is caused by sin, the best place for people with disabilities to live is institutions, or people with disabilities are less capable to contribute to society”.

The key contributing factors to these barriers are the knowledge, attitudes, and practice of key health and rehabilitation professionals, public servants, teachers, employers, employees, educators, counsellors, parents, peers, communities, etc. and persons with disabilities themselves. Understanding the attitudes of persons without disabilities towards persons with disabilities will help policy-makers to determine the nature of these negative attitudes and assess their extent, and to design appropriate interventions.
This research methodology was developed in consultation with key stakeholders working on disability issues, as well as with social scientists from the Tajik National University (TNU). It was informed by a literature review of policy, practice and projects for the well-being of persons with disabilities, with special focus on children and women. Further, consultations were held with the Directorate of Social Protection of the Ministry of Health and Social Protection of the Population, Handicap International, Mission East, NGO Zarshedabonu in Kulyab, Nur Centre in Khorog, and professors from the Department of Sociology of TNU.
Study design

In this study of KABPSN about persons with disabilities (especially towards children and women who are the most vulnerable) both quantitative and qualitative methodological approaches are proposed. This mixed methodological approach aims to collect and analyse in-depth data about the knowledge, attitudes, behaviour and practice of key stakeholders and community members in Tajikistan. The quantitative approach should be applied to conduct a survey (using a structured interview questionnaire) among community respondents aged 18 and above across gender both in rural and urban settings. For this purpose, 2,000 respondents need to be interviewed (see details in the Sampling section of this study design). The collected data will include, but will not be limited to, their demographic characteristics, educational levels, and KABPSN towards persons with disabilities.

For qualitative data collection, the study should cover the opinion and decision-makers at community level, such as service providers in day-care centres, teachers, primary healthcare workers, religious leaders, government officials and CwD, WwDs etc. In-depth Key Informant interviews should collect information on barriers that people with disabilities and their families at community level face in accessing healthcare, education, employment, social events, and information on their inclusion in the society in general.
Individual interviews or key informants interviews (KII) shall be conducted in selected clusters (out of 267 clusters in this study) where approximately 120 respondents will be interviewed through KII. The target population of the KII are opinion and decision-makers, such as local and national government officials, community leaders (heads of Mahala), religious leaders, caregivers (including mothers-in-law), service providers, teachers, school directors, healthcare workers, etc. This is because the decisions taken by these opinion and decision-makers affect the lives of persons with disabilities, particularly of children and women. The list of authorities for KII is provided at the end of this document.
Sample size

Sample size calculation was done by using Krejcie and Morgan formula which is commonly employed in the field of social sciences.

where:

X² – the table value of chi-square for one degree of freedom at the desired confidence level

N – the population size

P – the population proportion (assumed to be .50 since this would provide the maximum sample size)

d – the degree of accuracy expressed as a proportion (.05)
For a given population of 7,898,321 in Tajikistan in 2013¹, a sample size of 384 would be needed to represent a cross section of the population. Taking into consideration non-responses, invalid and missing cases, the sample size was rounded up to 400 and multiplied by 5 (five major administrative areas of Tajikistan: Dushanbe, GBAO, Khatlon, Sughd, and Region of Republican Subordination). The total sample for the survey is to include 2,000 individuals. During the survey/data collection, households can be used as a proxy for individuals.

The 2,000 sample segments were allocated by five regions by reflecting the proportion of the regional population in each area with the following formula:

Where:

n_i - is the samples size in strata (domain) i

n – is the sample size (in our case it is 2,000)

N – is the population
The final population by each strata (domain/region) is given in Table 4 below.
Sample Frame

The survey will focus on respondent aged 18 and above (males and females). In addtion, from every household the prospective consulting firm shall interview a child aged between 10 to 17, whenever possible. The sample size for this survey is 2,000 respondents. The sample size covers all the regions of the country in two-stage cluster sampling in order to reflect geographical and urban/rural diversity. The number of households to be interviewed in each cluster locality is calculated in a way that reflects the proportion of the population living in different regions and areas. Thus, for example, if 9% of the population lives in Dushanbe, 9% of the interviews from the selected proportion need to be conducted in Dushanbe. Through this method, it is hoped to approximate a nationally representative sample. Based on this selection, interviewers should be given a total number of individuals from a household that need to be contacted in each locality. The number of households in each locality should be selected through a process that is either random or is performed in a systematic way that eliminates any potential biases.

The estimated number of people to be interviewed is given in Appendix A. Table 1 represents the distribution of the population by region and settlement size. In accordance with the data in Table 1, the percentage of the distribution is presented in Table 2. The number of interviews required in accordance with the proportion of every region and settlement size (given in Table 2) is presented in Table 3. The overall number of Primary Sampling Units (PSU) is 267 units. The average number of interviews is given in Table 4. The more accurate allocation of interviews is given in Table 5.

The sample must be representative at national level in terms of gender. It is expected that the above-mentioned process should lead to roughly equal number of men and women being interviewed, as well as a child from each household. The gender of a child should also be roughly equal, but not statistically representative. The prospective consulting firm is required to keep the records of, and to monitor the number of men and women interviewed, and to outline a process to ensure adequate representation of gender, should they approve disproportionate representation of gender.

Data collection

The data collection shall be done through face-to-face interviews at respondent’s household of the key population. The exact timing of the interviews shall be decided in the most optimum way to reach the target population. Screening questions shall be made to identify the respondents in the household who belong to target population. If no suitable person is present for the interview, the interviewers will have to move to the next household, starting again from screening questions. The screening questions should be comprised of a maximum of 4 to 5 questions to determine the eligibility of the household to participate in the survey.

Proper statistical software should be used for data entry, for example, CSPro, as well as, data analysis computer software, such as, SPSS, STATA, SAS, etc. It is also advised to use computer software applications, such as, NVivo for analyzing the qualitative data acquired through individual interviews with key informants.
Components of the questionnaire

The questionnaire should consist of the following components:

1. 
   Socio-demographic characteristics, including, gender, age, number of children, etc.
   
2. 
   Education; job information;
   
3. 
   Knowledge section
   
4. 
   Attitude section
   
5. 
   Practice section
   
6. 
   Behaviour section
   
7. 
   Social Norms section
   

Variables

The prospective consulting firm is requested to be creative in terms of developing variables to measure knowledge, attitudes, behaviour and practice towards persons with disabilities, with special focus on children and women with disabilities. Different types of variable should be used to measure the KABPSN. Under Appendix B Table 1, some examples of variables are given. The questionnaire should also be developed in a way that captures the differences in attitudes towards males and females with disabilities.

Appendix A

Table 1. Distribution of population by region and urbanization level (Republic of Tajikistan)

		1	2	3	4	5	6
		1MM+	500K-999K	100K-499K	50K-99K	10K-49K	Under 9K	Total
1	Dushanbe		764 300					764 300
2	Sughd			167 300	110 584	1 845 700	191 549	2 315 133
3	Kurghonteppa zones				128 416	1 485 244	68 144	1 681 804
4	Kulob zones				98 500	827 821	168 463	1 094 784
5	RRS				68 700	1 597 374	258 205	1 924 279
6	MBAR					28 600	178 128	206 728
	Total	0	764 300	167 300	406 200	5 784 739	864 489	7 987 028

Table 2. Distribution of population by region and urbanization level in Republic of Tajikistan (%)

		1	2	3	4	5	6
		1MM+	500K-999K	100K-499K	50K-99K	10K-49K	Under 9K	Total
1	Dushanbe		9,6					9,6
2	Sughd			2,1	1,4	23,1	2,4	29,0
3	Kurghon-Teppa zones				1,6	18,6	0,9	21,1
4	Kulob zones				1,2	10,4	2,1	13,7
5	RRS				0,9	20,0	3,2	24,1
6	MBAR					0,4	2,2	2,6
	Total	0	9,6	2,1	5,1	72,4	10,8	100,0