Method of selection of households administered by long questionnaire and how to fill forms used for supervising enumeration areas (EAs)
For the 2007 population and Housing Census of Ethiopia two types of questionnaires, namely; short and long are used. The long questionnaire is a detailed one which will be administered to sample households. The short questionnaire with a subset of the long questionnaire items corresponding to basic demographic and social characteristics (population size, sex, age, religion, mother tongue, ethnic group, disability and orphanage) are administered to the remaining (non sample) households.
The rationale for administering long questionnaire in the census on a sample basis is to improve timeliness in data collection, processing and decrease costs associated with data collection and processing by reducing volume of the data and consequently improve quality of the data. In each EA of all regions a one in five (20 Percent) households are systematically selected for the administration of the long questionnaire.
A. selection of sample households administered by long questionnaire
For selecting sample households in each EA the sampling frame is form 1 which is used for listing housing units and households. The selection will be done using column 12 and 13 in section 2 of form 1. Before starting sample household selection, the enumerator needs to make sure that the listing is completed in each EA without omission. In addition, he should have to check that the running numbers in column 12, which are given for household, are done without omission and duplication. Since all hotel/hostels and collective quarters residents are administered by long questionnaire without selection, make a tick in column 14 corresponding to them to distinguish them from the others. Next, by using column 12 of form 1 select one household from each of the intervals 001 - 005, 006 - 010, 011 - 015...etc. To do this the enumerator will be given a "random start" number by his/her field supervisor to carry out the selection of households which lie between 1 - 5 including 1 and 5.
This random start number helps the enumerator to decide (Know) which households are going to be selected from his EA. The household with serial number in the list corresponding to the random start number is taken as the first sample household to receive the long questionnaire. The remaining sample households in the EA are determined by adding "five" consecutively on serial number of the current sample household. This process continued until the list of households is exhausted in each EA.
If the random start number 1 assigned to an enumerator by his/her supervisor the household corresponds to household serial number serial number 001 in column12 will be the first selected household in that EA. And also if another enumerator gets random start number 2, the household corresponds serial number 002 is the first selected household for long questionnaire. The same will apply for enumerators who get 3, 4 and 5 as a starting number.
To get the second sample household each enumerator should have to add 5 on the serial number of the first selected household and to get the third sample household they have to add 5 on the second selected household serial number and continue in that fashion to get the last sample household.
For example:- an enumerator whose assigned 1 as a starting number, households corresponds to serial numbers 001, 006, 011, 016, 021, 031...etc. in column 12 will be the sample households. Similarly, enumerators who get 2 as a starting number will select households with serial numbers 002, 007, 012, 017... etc. to fill the long questionnaire. Then a tick mark should be made in column 14.
While, doing the selection the enumerator should have to understand that all selected samples are regular households. Whenever, the sample lay on collective quarters or hotel/hostel skip it and take the next household. Remember all residents of collective quarters and Hotel/Hostels are administered by long questionnaire.
For example: - Suppose the starting number given to an enumerator is 2 and if on form 1 in column 12 serial number 002 is a collective quarter which is ticked on column 14, then leave that and take the household with serial number 003 as the first and to get the second selected household add 5 on the first. Continue this procedure until the list of households is exhausted.
B. Form for supervising enumeration area (EA).
This form used to check how many of short and long questionnaire are filled in correctly by the enumerator.
In addition, on this form identification number for the enumerator is available. Every enumerator should have to know and remember this number any time. This form is will be filled in after the enumerator completed the enumeration for every household using short and long questionnaire.
First, the enumerator records the identification particulars of the EA on the left side of this form by taking it from the EA map. Then, on the right side under the box, which says "number of completed questionnaires", the enumerator records the number of short and long questionnaires he completely filled in the information.
After making every correction, the enumerator will write his name put his signature and date on the form. In addition, the supervisor will do the same after checking the correctness of the information on the questionnaires. Then the form should send to head office with completed short and long questionnaires.