The primary objectives of the Punjab MICS 2007-08 are:
- To update and compare the results of Punjab MICS 2003??04 with finding in Punjab MICS 2007-08 at the district and tehsil levels;
- To establish benchmarks indicators which were not included in Punjab MICS 2003-04 but are included in the Millennium Development Goals (MDGs);
- To highlight inter-district disparities on the basis of evidence in order to address them through appropriate district and tehsil-level social sector planning efforts by the provincial government;
- To develop a strong advocacy tool for action on poor social service delivery; To provide information on the situation of children and women and assist the government in establishing child-focused benchmarking to report on the MDGs and the government's long term plans, and to measure progress;
- To build the capacity of government institutions by encouraging their active involvement in all phases of the survey.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
The survey covers the following themes:
- Characteristics of households and respondents
- Education: literacy; pre-primary and secondary school participation;distance from educational facility
- Environment: water and sanitation, source of drinking water
- Adult health and health care: reported chronic cough; TB and hepatitis, care provided by lady health worker;physical access to health facility
- Child mortality
- Child Mortality: nutritional status; breastfeeding; salt iodisation
- Child health: oral rehydration therapy; care seeking for suspected pneumonia; prevalence of recent illness; solid fuels
- Child protection: birth registration; child labour; child disability;
- Reproductive health: children ever born; fertility; contraception; unwilling pregnancy; antenatal care; assistance at delivery; postnatal care; knowledge of preventing HIV/AIDS; attitudes towards people living with HIV/AIDS
- Socioeconomic Development: unemployment; housing and assets; remittances and cash donations; social benefits, subsidies and family support programmes
The universe consists of all the households and their members in all urban and rural areas of Punjab as defined for the 1998 Census of Population and Housing (CPH) and subsequent changes made by the provincial government. Areas involving military installations and the homeless are excluded from the scope of the survey. The province of Punjab is administratively divided into 9 divisions, 35 districts and 143 tehsils / towns.
Producers and sponsors
Bureau of Statistics, Planning and Development Department
Government of Punjab, Lahore
Manar E. Abdel-Rahman, PhD
M/s Eycon Pvt. Limited
Data Management Consultants
The sample for the Punjab MICS 2007-08 was designed by the Federal Bureau of Statistics (FBS), Government of Pakistan to provide estimates of socio-economic indicators at the provincial level for nine divisions, 35 districts, 143 tehsils/towns, major cities, other urban and rural areas (Table SD.1 in Appendix A of the survey report available as external resources). The sample design was reviewed for adequacy and soundness by an international consultant engaged by UNICEF. Sample size summarised by district is presented in Table SD.1A. The sample was selected in two stages. Within each of the 273 sampling domains1, enumeration areas (enumeration blocks in urban areas or village/mouzas/dehs in rural areas) were selected with Probability Proportional to Sizes. Household listing was carried out within each randomly selected enumeration areas and a systematic sample of 12 households in urban areas and 16 households in rural areas was randomly drawn. The total sample size for the survey was 91,280 households. The sample was not self-weighting and sample weights were used to report results. A more detailed description of the sample design can be found in Appendix A.
Deviations from the Sample Design
Sometimes it happens that for a reason outside our control (war, flooding, security reasons, etc) an EB cannot be enumerated. If such were the case, it is recommended to have available procedures that might allow us to replace one sample EB with another. If this needs to be done, follow the following recommendations:
1. Within the domain of estimation and the socioeconomic substratum to which the EB that has to be replaced belongs, select a random number between 1 and the cumulative measure of size within the domain and socioeconomic substratum.
2. Locate the interval within which the selected random number fell. If the selected EB is already in sample, don't use it.
3. Repeat step 1 again. Locate the interval within which the selected random number fell. If the EB was not previously selected, this EB will be now in sample and it will replace the EB that could not be initially utilised.
4. All the procedures mentioned in this report must be followed for the new EB selected to replace another EB. That is to say, if the number of HHs is very large, the EB must be segmented. One or more segments must be chosen at random if the circumstances so require it. The new EB must be listed, just as the EB that was
replaced was listed.
In order to expand the data from the Punjab MICS 2007–08 to the provincial or domain levels (tehsils), it is necessary to apply a weight (expansion factor) to the data from each survey questionnaire. The basic weight for a sample household would be equal to the inverse of its probability of selection.More details on weighting are given in Section II, Appendix A of the report available under external resources.
Dates of Data Collection
Data Collection Mode
Data Collection Notes
Due to the large sample size and to ease management of the large fieldwork operation while assuring quality and proper supervision, the province was divided into ten regions of 3-5 districts each: Bahawalpur, DG Khan, Faisalabad, Gujranwala, Lahore I, Lahore II, Multan, Rawalpindi, Sahiwal and Sargodha. The BOS assigned one senior officer as Regional Supervisor who was responsible for all aspects of fieldwork in his or her region. Household listing began in October 2007. The Federal Bureau of Statistics (FBS) trained 300 listers taken by BOS from line departments and academia, and supervised the process through its regional offices.
A 5-day training of 22 master trainers was conducted in Lahore during November 2007. Master trainers comprised of technical staff from the BOS, female doctors from the health department and professional trainers engaged by UNICEF. The training was conducted by a private sector consultancy firm, Employment through Creativity Integration Pvt Ltd. A manual was developed training interviewers and supervisory staff. Field enumeration staff were trained by master trainers at the regional level through 24 sessions spread over a week each in December 2007.
Each session was standardised in terms of duration, content, material and instructional methodology and provided adequate opportunities for hands-on practice and closely supervised field work. Standardised instruction manuals for team supervisors and interviewers were developed and used during training and fieldwork. Each training session concluded with a standard evaluation of the participants.
A total of 86 team supervisors led the interviewer teams. Each team comprised one male and two female interviewers. The male interviewers were responsible for locating sampled households, introductions, administering the household modules and performing water testing. The female interviewers administered the women and children modules to eligible women and mothers or caretakers of children. They also performed the anthropometric measurements on children. Depending on the size of the household, administration of each questionnaire took about 35-45 minutes.
Each team supervisor supported two teams in urban or three teams in rural areas. Supervisors ensured that the visited household was one identified through the sampling process. They were also responsible for completeness and quality of work, keeping to schedules, and arranging transport and accommodation. Team supervisors were the primary link between regional supervisors and interviewers, and reported on their teams' progress and performance regularly. Teams were instructed to complete one cluster a day. All cluster and control sheets were delivered by team supervisors to their regional supervisors who forwarded them to the BOS headquarters in Lahore after two days to allow the completion of the water test.
To establish the credibility of the survey, the University of the Punjab's Social Work Department was engaged as an independent monitoring organisation and was involved in the survey process from the beginning. This third party monitoring was intended to support, assess and report on the quality of all aspects of the survey work. The monitors covered questionnaire design, training, listing, fieldwork and back-checking of filled-in questionnaires, data management and report writing. The third party monitoring report is included in Appendix D.
Team supervisors were largely taken from the BOS, Crop Reporting Service (CRS), Agriculture, Health and Social Welfare departments. The interviewers were engaged from the BOS and other government line departments including Population Welfare, Health and CRS and included university graduates through the Gender Reform Action Plan and NGOs. In total, 669 interviewers, 86 team supervisors and 10 regional supervisors were involved in data collection. Thus 765 field enumeration staff participated in fieldwork for the survey from 12 December 2007 to 10 April 2008.
Questionnaires for the Punjab MICS 2007-08 were based on MICS-3 model questionnaires with some modifications and additions to accommodate additional indicators selected by the Government of the Punjab. A household questionnaire was administered in each sampled household, with three major sections: information on all household members, individual questionnaires for each ever-married woman aged 15-49 years, and a children's section administered to mother or caretakers of children under 5 years of age. The questionnaires included the following modules:
- Household Questionnaire
o Household Listing
o Income and Employment
o Water and Sanitation
o Child Labour
o Maternal Mortality
o Household Characteristics
o Salt Iodisation
o Drinking Water Testing
- Questionnaire for Individual Woman
o Child Mortality
o Maternal and Newborn Health
- Questionnaire for Children Under 5
o Birth Registration
o Vitamin A
o Care during Illness
o BCG Immunisation
The MICS-3 model questionnaires were revised and customised to suit local conditions and translated from English into the Urdu language. The Urdu version of the questionnaire was pretested in November 2007 in the urban area of Faisalabad and the rural area of Kasur on a set of households that was not included in the survey sample. Questionnaires were modified based on the results of the pre-test.
Questionnaire responses received a preliminary edit on the spot by team supervisors. At the end of the day, the team supervisor and interviewers reviewed each questionnaire and filled in the cluster control sheet. The regional supervisors checked randomly selected questionnaires. Data editors deployed by the data management organisation conducted a final edit and check before data entry.
Data handling and processing was managed by an experienced private sector agency which was responsible for survey data management and processing, including pre-entry editing, data entry, verification and consistency checks, preliminary tabulations and frequencies, support of the data analysis process and preparation of final data for dissemination in electronic form. In total, 48 data editors, 78 data entry operators and eight data supervisors and quality control staff took part in the data processing operation. They were trained in four batches during May 2008. The two trainings for editors were conducted by Punjab MICS master trainers following the curriculum used in training interviewers.
For better quality control, all questionnaires were double entered and internal consistency checks were performed. The data management operation site was housed within the BOS premises to ensure close monitoring. A short-term international consultant was engaged by UNICEF to review the technical aspects of data entry operation during June 2008 who recommend improvements in the data management process.
Data processing began on 10 May and concluded on 15 September 2008. Procedures and standard programs developed under the global MICS-3 project and adapted to the Punjab questionnaires were used. Data were entered and processed using the CSPro software after customising standard data entry programs, and were analysed using the Statistical Package for Social Sciences (SPSS) software program after customising the model syntax and tabulation plans. Syntax and tabulation plans were prepared for the additional indicators included in the survey. Table numbers used in this report refer to the same numbers used in the global MICS-3 tabulation plan. Each table number is prefixed by two letters denoting the section to which the table belongs. Generally numbering of figures reflects the table numbers from which the data used in the figures are obtained.
Data cleaning took place at three stages as follows:
- Desk editing of questionnaires by the DMF;
- Sample editing of questionnaires by the third party editors;
- Reference to original questionnaires in double data entry system;
- Data analysis.
At the desk editing stage, the DMF editors went through the questionnaires to ensure internal consistency of data. At this stage, most of the missing values were imputed on the basis of undeniable evidence within the questionnaire. The third party editors pointed out some leftover data editing issues on sample basis, most of which could be attended through computer-based logical checks. The DMF worked on them and improved the data accordingly.While comparing two datasets created through independent data entries, most data entry errors were addressed. In this process the two datasets were compared and differences were removed by referring to original questionnaires. The result was a third dataset which was reasonably clean of data entry errors.
Estimates of Sampling Error
The sample of respondents selected in the Punjab MICS is only one of the samples that could have been selected from the same population using the same design and size. Each of these samples would yield results that differ somewhat from the results of the actual sample selected. Sampling not known exactly, but can be estimated statistically from the survey results.
The following sampling error measures are presented in this appendix for each of the selected
o Standard error (se): Sampling errors are usually measured in terms of standard errors for particular indicators (means, proportions etc). Standard error is the square root of the variance. The Taylor linearisation method is used for the estimation of standard errors.
o Coefficient of variation (se/r) is the ratio of the standard error to the value of the indicator
o Design effect (deff) is the ratio of the actual variance of an indicator, under the sampling method used in the survey, to the variance calculated under the assumption of simple random sampling. The square root of the design effect (deft) is used to show the efficiency of the sample design. A deft value of 1.0 indicates that the sample design is as efficient as a simple random sample, while a deft value above 1.0 indicates the increase in the standard error due to the use of a more complex sample design.
o Confidence limits are calculated to show the interval within which the true value for the population can be reasonably assumed to fall. For any given statistic calculated from the survey, the value of that statistics will fall within a range of plus or minus two times the standard error (p + 2.se or p - 2.se) of the statistic in 95 per cent of all possible samples of identical size and design.
For the calculation of sampling errors from Punjab MICS data, SPSS Version 13 Complex Samples module has been used. The results are shown in the tables that follow. In addition to the sampling error measures described above, the tables also include weighted and unweighted counts of denominators for each indicator.
Sampling errors are calculated for indicators of primary interest, for the Punjab total and areas of residence. Four of the selected indicators are based on households, 13 are based on household members, 7 are based on children under-five, and 5 are based on women. All indicators presented here are in the form of proportions. in Appendix B of the survey report, Table SE.1 shows the list of indicators for which sampling errors are calculated, including the base population (denominator) for each indicator. Tables SE.2 to SE.7 show the calculated sampling errors.
Bureau of Statistics
Planning & Development Department, Government of the Punjab
Before being granted access to the dataset, all users have to formally agree:
1. To make no copies of any files or portions of files to which s/he is granted access except those authorized by the data depositor.
2. Not to use any technique in an attempt to learn the identity of any person, establishment, or sampling unit not identified on public use data files.
3. To hold in strictest confidence the identification of any establishment or individual that may be inadvertently revealed in any documents or discussion, or analysis. Such inadvertent identification revealed in her/his analysis will be immediately brought to the attention of the data depositor.
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Bureau of Statistics
Planning and Development Department, Government of the Punjab
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.