Statistical Methods, Standards and Guidelines 52
Questionnaires
Four questionnaires are used for TDHS-MIS: the Household Questionnaire, the Woman’s
Questionnaire, the Man’s Questionnaire and the Biomarker Questionnaire. The Household
Questionnaire is used to list all the usual members and visitors in the selected households. Basic
demographic information are collected on the characteristics of each person listed, including age,
sex, education, and relationship to the head of the household. The Household Questionnaire also
collects information on characteristics of the household’s dwelling unit, such as source of drinking
water, type of toilet facilities, ownership of durable goods and assets and ownership and use of
mosquito nets. The Household Questionnaire saves the purpose of identifying women and men
who are eligible for the individual interview. The Woman’s Questionnaire is used to collect
information from all women age 15-49. The Man’s Questionnaire is administered to all men age
15-49 living in every third household (subsample) in the TDHS-MIS sample. The Man’s
Questionnaire collects much of the same information available in the Woman’s Questionnaire, but
it does not contain a detailed reproductive history or questions on maternal and child health. The
Biomarker Questionnaire is used to record anthropometric measurements (height and weight) for
children under age 5 and women age 15-49, record anaemia test results for children age 6-59
months and women age 15-49; record malaria rapid test results for children age 6-59 months;
document responses to a request for blood samples among children age 6-59 months, to be tested
later for malaria using microscopy in the laboratory; and document responses to request for a
household salt sample and a urine sample among women age 15-49, to be tested later for iodine in
the laboratory.
Biomarker Testing The TDHS-MIS consists of four biomarker testing; Anthropometry (Height and Weight Measurement), Anaemia Testing, Malaria Testing and Iodine Testing.
Pre-test: All elements of the survey are to be pretested prior to the main survey.
Training of Field Staff:
The training is conducted following the DHS training procedures, including classroom
presentations, mock interviews, tests and field practice (which combine interviews and collection
of biomarkers).
Statistical Methods, Standards and Guidelines 53
Fieldwork: Field work is done in form of teams, of which each consists of four female interviewers, male interviewer, a supervisor, a field editor, and a driver. The field editor and supervisor are responsible for reviewing all questionnaires for quality and consistency check before the team’s departure from the cluster.
Data Processing: In the TDHS-MIS the data entry is done concurrently with data collection in the field. Filled-in paper questionnaires in the field are edited, and checked by both the field editor and the supervisor; the data is then entered into a tablet equipped with a data entry programme (CSPro Software). Completed questionnaires are then sent to NBS headquarters, where they are entered for the second time and edited by data processing personnel. Processing the data concurrently with data collection allows for regular monitoring of team performance and data quality. Field check tables are generated regularly during data processing to check various data quality parameters. As a result, feedback is given on a regular basis, encouraging teams to continue in areas of good performance and to correct areas in need of improvement.
Quality Control: Quality control teams periodically visit teams in the field to check their work and re-interview some households (spot checks).
Analysis and Report writing: After all completed questionnaires have been checked, entered and finalized and analysis completed, at least two reports have to be prepared: Preliminary report and final report. Preliminary report has to be brief and consists primarily of key indicators. The final report is more detailed.
Dissemination and Utilization of Results A National dissemination seminar is conducted after completion of report writing. A series of zonal and regional level dissemination seminars are also done if funds are available.
4.2.3 Reference to applicable standard classifications The TDHS-MIS follow the International reference classifications such as Tanzania Standard Classification of Occupations (TASCO).
Statistical Methods, Standards and Guidelines 54
4.3
TANZANIA SERVICE PROVISION ASSESSMENT (TSPA)
Tanzania Service Provision Assessment (TSPA) survey is a sample survey of health facilities
which provides information about the overall service environment, resources, practices and
functioning of components of the health system that may affect the quality of health services in
Tanzania.
4.3.1 The objectives and uses of TSPA
The general objective of the Tanzania Service Provision Assessment (TSPA) is to collect
information on the delivery of health care services in Tanzania and to examine the readiness of
facilities for provision of quality health services in the country. These services are in the areas of
child health, maternal and new born care, family planning, sexually transmitted infections, HIV
and AIDS, tuberculosis, malaria, and chronic diseases.
Uses of TSPA:
The findings of the survey are used to provide snapshot information on overall availability of
different health related services. In addition, it assesses the readiness of health facilities to operate
effectively and efficiently in order to provide quality services to clients. Findings from this
assessment are used to design and improve interventions and services geared at ensuring the
population can assess the necessary quality services with minimum delay and maximum
satisfaction.
4.3.2 Methodology of data production
Sample Design
The sample for the TSPA is designed to provide nationally representative results by facility type
and managing authority and regionally representative results for the regions in Tanzania Mainland
and Zanzibar. A sample of health facilities selected to participate in the TSPA survey is drawn
from a master list of health facilities that consists of verified (active) health facilities in Tanzania.
The master list contains the list of all hospitals, health centres, dispensaries, and clinics. These
facilities are managed by the government, private-for-profit, parastatal, and faith-based entities.
The sample design for TSPA is a stratified random sample of health facilities selected with equal probability under systematic sampling. The stratification is achieved by separating the health
Statistical Methods, Standards and Guidelines 55
facilities by facility type within each region. The sample allocation features a power allocation across regions in order to achieve comparable survey precision across regions.
Questionnaires
Four questionnaires are used for TSPA: the Facility Inventory Questionnaire, the Health Provider
Interview Questionnaire, the Observation Protocol Questionnaires for antenatal care (ANC),
Family Planning (FP), services for Sick Children (SC) and the Exit Interview Questionnaires for
ANC, FP and SC.
Data Collection Approaches The inventory questionnaires are loaded onto tablet computers, which are used during interviews to ask questions and also record responses (computer assisted personal interviewing–CAPI). All other types of questionnaires are paper based, but responses are entered into computers and edited in the field (computer assisted field entry–CAFE).
Pre-test: All elements of the survey are to be pretested prior to the main survey.
Training of Field Staff:
The training is conducted following the DHS training procedures, including classroom lectures
and discussions, practical demonstrations, mock interviews, role-plays, and field practices.
Fieldwork: After the training, the data collection personnel are divided into teams. Each team consists of a team leader, 3 interviewers and a driver.
Data Processing and Questionnaire Management: After completing data collection in each facility, the interviewers review the paper questionnaires (Health Provider Interview, Exit Interview and Observation) and the Inventory data that have been collected directly onto the tablet computer before handing the questionnaires and electronic data over to the team leader, who reviews them a second time. The paper questionnaires are then entered into the second tablet computer. Once data collection and all data entry are completed in a facility, the team leader conducts consistency and structural checks on the data to identify any errors or missing information. When a team is satisfied that data collection and entry are complete
Statistical Methods, Standards and Guidelines 56
for the facility, the team sends the data to the NBS headquarters via the Internet, using Internet File Streaming System (IFSS). Completed paper questionnaires from the field are periodically picked up by the quality control teams and taken to the NBS headquarters. At the NBS headquarters, the data from the completed paper questionnaires from the field are entered twice (100 percent verification). The concurrent processing of the data has a distinct advantage for data quality.
Quality Control: Quality control teams periodically visit teams in the field to check their work and re-interview some facilities (spot checks).
Analysis and Report: After all completed questionnaires have been checked, entered and finalized and analysis completed, at least two reports have to be prepared: Preliminary report and final report. Preliminary report has to be brief and consists primarily of key indicators. The final report is more detailed.
Dissemination and Utilization of Results A national dissemination seminar is conducted after completion of report writing. A series of zonal and regional level dissemination seminars are also done if funds are available.
4.3.3 Reference to applicable standard classifications The TSPA follow the International reference classifications such as Tanzania Standard Classification of Occupations (TASCO).
4.4
TANZANIA DISABILITY SURVEY
The Tanzania Disability Survey is the household-based and comprehensive survey of its kind
carried out in Tanzania. The survey provides information on the prevalence of disability in
Tanzania.
4.4.1 The objective and uses of TDS:
The objective of the Tanzania Disability Survey (TDS) is to determine the prevalence of disability
in the country and to determine living conditions among people with activity limitations.
Statistical Methods, Standards and Guidelines 57
Use of TDS The findings of the survey are used as a contribution to the improvement of the living conditions among people with activity limitations in Tanzania; the findings also initiate a discussion on the concepts and understanding of “disability” and monitor the impact of government policies, programmes and donor support on the well being of the population with activity limitations.
4.4.2 Methodology of data production
Sampling Design
The survey covers the whole of the United Republic of Tanzania and produces estimates at
regional level. Information is collected from all selected households and individuals with and
without disability (measured as activity limitations).
The primary sampling unit for the survey is the census enumeration area (EA) and the ultimate
sampling unit are the individual household members. Disability survey utilizes a three-stage
systematic stratified random sampling design, involving clusters (EAs), households and individual
household members.
Sampling Frame The sampling frame of clusters to be used is the list of all enumeration areas (EAs) generated during the Population and Housing Census. EAs in each region are listed following the census coding system and a target sample is selected using probability proportional to size. The EA maps and other administrative information are used to identify the boundaries and features of the selected EAs. For households, the sampling frame should be the list of households (heads) constructed for each selected EA. To ensure a random scattered sample, the listing of households should be done in a serpentine manner from one end of the EA to another end.
Sample Size Determination A total number of clusters must be selected. The targeted sample per cluster should be determined. The selection of EAs should follow the Probability Proportional to Size (PPS) sampling while the selection of households and individuals follow a simple random sampling procedure. The random spread of households is necessary for achieving a non-clustered sample.
Statistical Methods, Standards and Guidelines 58
Questionnaire Disability survey uses three types of questionnaires. The first questionnaire is Household Questionnaire, which attempts to collect basic socio-economic conditions of usual household members and visitors. The questionnaire also includes screening questions which are used to determine persons with disabilities.
The second questionnaire is Adult Questionnaire which is used to collect information from all person aged 15 years and above who were identified in the household questionnaire as having some form of disability.
The Children Questionnaire is the third instrument which is used to collect information from all children identified as having disabilities but collect more or less the same information in the adult questionnaire.
Training of Field Staff The training involves both theory and practise. Organisations of People with Disability are represented by participants who have a responsibility of clarifying terms used in the questionnaires. Besides participating in training representatives from Organisations of People with Disabilities is also part of supervisory teams.
Field work The field work includes map reading and listing exercise, the supervisors and enumerators must be trained on map reading and listing. Supervisors should ensure that all households in the EA are listed according to the given instructions and EA map. Supervisors are also responsible for selecting households to be interviewed.
Data Collection Data collection is conducted in teams, each team consist of supervisor, enumerators and a driver. Supervisors are responsible for the overall administrative work in the field including checking the quality of the questionnaires before departing from the cluster.